Working with Agentic AI: Systems Engineer
1. Working with Agentic AI: Systems Engineer
How Human Skills and Agentic Skills Dynamically Combine Across Systems Engineering Responsibilities
1.1. Systems Engineering as Professional Work
A Systems Engineer works across the technical and organizational boundaries of complex engineered systems to help ensure that system elements, interfaces, requirements, designs, processes, and lifecycle activities function together as a coherent whole. The role involves understanding stakeholder needs, defining and analyzing system requirements, developing and evaluating system architectures, modeling system behavior and relationships, managing interfaces and dependencies, evaluating design alternatives, identifying technical risks, coordinating system integration, supporting verification and validation, maintaining requirements traceability, managing technical changes, investigating system-level problems, and coordinating engineering information across different technical disciplines. Systems Engineers may also contribute to technical planning, design reviews, configuration management, system transition, operation and maintenance, quality and compliance activities, and continuous improvement of engineering processes. Their responsibilities therefore extend beyond designing individual components: they continuously connect stakeholder needs, system requirements, architecture, hardware, software, data, interfaces, constraints, risks, verification evidence, operational conditions, and lifecycle considerations while applying systems thinking and professional judgment to determine how individual elements and engineering decisions affect the behavior and effectiveness of the system as a whole.
1.2. Working with Agentic AI
Agentic AI can participate across Systems Engineering activities by applying the comprehension, reasoning, comparison, generation, coding, evaluation, classification, and multimodal capabilities of increasingly advanced large language models and other foundation models together with engineering context, technical instructions, external knowledge sources, specialized tools, software environments, and reusable or dynamically constructed agentic skills. Within a system lifecycle, an AI agent can process stakeholder needs and technical documentation, extract and organize system requirements, retrieve relevant engineering knowledge, analyze dependencies and interfaces, support functional and architectural decomposition, generate and compare preliminary design alternatives, assist with system modeling and technical analysis, maintain requirements traceability, examine risks and change impacts, prepare verification and validation artifacts, analyze integration or test evidence, identify inconsistencies across engineering artifacts, and consolidate information distributed among different engineering disciplines. These activities can also be organized into goal-directed workflows in which the agent decomposes an engineering task into subtasks, selects appropriate information and tools, performs intermediate analyses, evaluates partial results, and adjusts subsequent actions as requirements, designs, test findings, risks, or other system conditions change. Their practical value does not depend on transferring an entire Systems Engineering responsibility to AI. Instead, agentic skills can participate selectively across individual tasks and subtasks, while the Systems Engineer provides systems-level interpretation, engineering judgment, interdisciplinary coordination, validation, and decision-making where the technical context, consequences, uncertainty, or authority of the work requires them.
1.3. Human Skills and Agentic Skills
Human skills in Systems Engineering refer to the professional capabilities that a Systems Engineer applies to understand and manage relationships across stakeholder needs, system requirements, functions, architectures, interfaces, system elements, technical disciplines, risks, verification evidence, operational conditions, and lifecycle decisions. They include systems thinking, engineering knowledge, abstraction, experiential judgment, technical reasoning, trade-off evaluation, interdisciplinary communication, risk judgment, and the ability to recognize how a decision affecting one part of a system may produce consequences elsewhere. Human skills are particularly important when requirements or evidence are incomplete, engineering alternatives involve competing objectives, different disciplines have conflicting priorities, system-level consequences cannot be determined from individual components alone, or technical decisions involve significant safety, security, operational, financial, or organizational consequences. Systems Engineers also perform analytical activities that may overlap with AI capabilities, including requirements analysis, comparison of alternatives, modeling, problem investigation, evidence evaluation, and technical explanation. Human capability in Systems Engineering should therefore not be defined simply by separating human analytical work from activities that AI can also perform.
Agentic skills in Systems Engineering are task-oriented capabilities through which an AI agent can participate in the analysis, development, integration, verification, and lifecycle management of engineered systems. Depending on the available models, engineering information, tools, and authorization, these skills may extract and classify requirements, analyze requirement quality and dependencies, construct or examine traceability relationships, retrieve engineering standards and technical knowledge, compare architecture or design alternatives, analyze interfaces, generate preliminary models or analytical code, examine risk information, assess the potential propagation of technical changes, organize integration and test evidence, identify inconsistencies among engineering artifacts, support verification coverage analysis, summarize technical reviews, and investigate patterns across logs, test results, configuration histories, or operational records. Agentic skills can draw simultaneously on model capabilities and system-specific information contained in requirements repositories, architecture descriptions, interface specifications, models, configuration records, engineering databases, test systems, technical standards, and other authorized sources. Their usefulness consequently changes with the particular system, lifecycle stage, engineering tool environment, information quality, and reliability required for the task.
The important distinction in Systems Engineering is therefore not that requirements analysis, modeling, technical reasoning, risk analysis, verification, or other cognitive activities inherently belong either to the Systems Engineer or to AI. Both human and agentic capabilities may contribute to many of these activities, but their suitability can differ substantially according to the engineering situation. An agentic skill may efficiently analyze thousands of requirements for possible inconsistencies or traceability gaps, while a Systems Engineer may need to determine whether a flagged relationship is technically meaningful; AI may generate and compare architectural alternatives, while the engineer evaluates whether their assumptions and system-level trade-offs are acceptable; and AI may analyze integration or verification evidence, while consequential conclusions remain dependent on professional interpretation and engineering authority. The practical question is therefore which available human or agentic skill, or combination of skills, should perform a particular Systems Engineering task or subtask under the current system, lifecycle, technical, and organizational conditions.
1.4. Dynamic Allocation of Tasks and Subtasks
A professional responsibility such as requirements engineering, system architecture development, integration, or verification and validation is usually composed of multiple tasks and subtasks rather than being a single indivisible activity. These subtasks may require different combinations of comprehension, information retrieval, technical analysis, modeling, engineering judgment, communication, generation, evaluation, decision-making, and execution. The participation of human skills and agentic skills can therefore vary substantially within the same responsibility.
The allocation of work should be understood as dynamic rather than fixed. A subtask may initially be performed by an agentic skill, transferred to the Systems Engineer when system-level judgment is required, returned to AI for additional analysis after new information is obtained, and later validated by the Systems Engineer before the result is incorporated into an engineering decision or system artifact. Other subtasks may be performed independently by the professional or independently by an AI agent when their capabilities and authority are sufficient. Allocation can change according to task complexity, technical consequences, available information, model capability, tool availability, output reliability, system criticality, engineering standards, organizational policies, intermediate results, and changing lifecycle conditions.
A useful way to understand this structure is:
Occupation → Responsibility → Task → Subtask → Human and/or Agentic Skills → Evaluation → Possible Reallocation
The purpose of integrating agentic skills is therefore not to eliminate traditional professional skills. It is to expand the set of capabilities available for performing work and to enable the professional to determine, continuously and contextually, how human and agentic capabilities can be combined most effectively.
1.5. Systems Engineering Responsibilities
1.5.1. Stakeholder Needs and System Requirements
Stakeholder needs and system requirements establish what a system is expected to accomplish, the conditions under which it must operate, the qualities it must exhibit, and the constraints within which it must be developed and used. Human skills are particularly important when communicating with stakeholders, interpreting operational context, distinguishing underlying needs from proposed solutions, recognizing conflicting expectations, determining system boundaries, and deciding which concerns require deeper investigation. Agentic skills can process stakeholder discussions, operational documents, existing specifications, technical records, standards, and other authorized sources to extract candidate needs, organize them by topic, identify relationships, detect ambiguity or inconsistency, and generate possible clarification questions. The Systems Engineer may conduct initial stakeholder analysis, AI may structure and examine the resulting information, and identified gaps may lead to additional stakeholder interaction and subsequent agentic analysis.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Engage stakeholders, interpret operational context, distinguish needs from proposed solutions, identify conflicting expectations, and validate actual system needs. | Process discussions and documentation, extract and classify needs, identify relationships and ambiguities, summarize evidence, and generate possible clarification questions. | Human interactions generate information for agentic analysis; AI-detected gaps can guide further stakeholder inquiry; new information can subsequently refine system needs and requirements. |
1.5.2. Requirements Analysis and Management
Requirements analysis and management transforms stakeholder needs into clear, feasible, consistent, traceable, and verifiable system requirements and maintains them as the system evolves. Human skills contribute engineering judgment, contextual understanding, technical feasibility assessment, prioritization, and the ability to determine whether requirements correctly represent intended system behavior and constraints. Agentic skills can structure large requirement sets, normalize terminology, classify requirements, identify duplication or conflicts, detect ambiguous expressions, analyze dependencies, generate preliminary derived requirements, and compare changed versions. Requirements may therefore move repeatedly between agentic analysis and professional validation as design decisions, risks, stakeholder expectations, or system conditions change.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Interpret requirement intent, assess feasibility, resolve ambiguity and conflicts, establish priorities, and validate requirements against stakeholder needs. | Structure and classify requirements, identify duplication or inconsistency, analyze dependencies, compare revisions, and generate candidate derived requirements. | AI can analyze requirements at scale; human corrections and engineering decisions can change requirements; revised sets can then undergo repeated agentic analysis and checking. |
1.5.3. System Architecture and Design
System architecture and design defines the fundamental organization of a system, including its major elements, functions, relationships, interfaces, boundaries, allocations, and design principles. Human skills contribute systems thinking, engineering experience, abstraction, creativity, trade-off judgment, and understanding of technical and lifecycle constraints. Agentic skills can retrieve relevant engineering information, map requirements to architectural elements, generate candidate decompositions, identify possible architectural patterns, compare alternatives against specified criteria, analyze dependencies, and detect assumptions or inconsistencies requiring validation. Rather than simply generating one architecture, agentic reasoning can help explore a broader design space while the Systems Engineer evaluates alternatives and introduces system-level considerations that may not be adequately represented in the initial information.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Apply systems thinking, engineering experience, abstraction, and professional judgment to develop and evaluate technically coherent architectures. | Retrieve engineering knowledge, map requirements to system elements, generate candidate structures, compare alternatives, and identify dependencies, assumptions, or inconsistencies. | AI can explore preliminary architectural alternatives; human evaluation can change constraints or design priorities; AI can then regenerate or analyze revised architectures. |
1.5.4. System Modeling and Analysis
System modeling and analysis uses logical, functional, behavioral, physical, mathematical, computational, or other representations to understand how a system and its elements behave under different conditions. Human skills are important for determining what should be modeled, selecting appropriate levels of abstraction, interpreting model assumptions, recognizing limitations, and deciding whether analytical results are meaningful for engineering decisions. Agentic skills can assist with constructing preliminary models, generating model elements or code, retrieving parameter information, organizing model relationships, executing authorized analytical tools, comparing scenarios, summarizing results, and identifying anomalies. Human interpretation can then determine whether the model adequately represents the intended system and whether additional modeling or analysis is necessary.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Determine modeling objectives, select appropriate abstractions, evaluate assumptions, interpret results, and decide their engineering significance. | Generate preliminary models or analytical code, organize relationships and parameters, execute available tools, compare scenarios, and summarize results. | Human-defined questions guide agentic modeling; AI-generated results can reveal unexpected system behavior; professional interpretation can trigger model refinement or additional analysis. |
1.5.5. Interface Definition and Management
Interface definition and management identifies and controls the interactions among system elements and between the system and external systems, users, environments, or supporting infrastructure. Interfaces may involve physical connections, software interactions, data exchanges, communication protocols, energy, materials, timing, control signals, or operational responsibilities. Human skills are important for coordinating different engineering disciplines, resolving interface disagreements, understanding system-level consequences, and determining whether interface definitions are technically sufficient. Agentic skills can extract interface information from requirements and design documents, organize interface relationships, identify inconsistencies, compare definitions across artifacts, detect missing information, and maintain structured interface records as designs change.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Coordinate interface decisions across disciplines, resolve conflicts, evaluate system-level implications, and validate interface definitions. | Extract and structure interface information, compare definitions, identify inconsistencies or missing information, and maintain interface relationships across artifacts. | AI can continuously analyze interface information; human resolution of conflicts can update interface definitions; revised information can trigger renewed consistency analysis. |
1.5.6. System and Decision Analysis
System and decision analysis evaluates alternative technical approaches, architectures, technologies, configurations, or implementation strategies to support engineering decisions. Human skills contribute judgment about evaluation criteria, system priorities, uncertainty, trade-offs, and consequences that may not be completely represented quantitatively. Agentic skills can organize decision criteria, retrieve supporting information, generate alternatives, perform structured comparisons, analyze trade-offs, generate scenario analyses, and identify assumptions influencing the result. The Systems Engineer can evaluate the analysis, modify criteria or weights, introduce additional constraints, and delegate renewed analysis back to the agent.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Define meaningful decision criteria, interpret trade-offs and uncertainty, evaluate system-level consequences, and take responsibility for engineering decisions. | Retrieve evidence, generate alternatives, structure criteria, compare options, analyze scenarios, and identify assumptions or dependencies. | AI can perform broad alternative analysis; human judgment can modify criteria or constraints; revised conditions can trigger repeated agentic comparison and evaluation. |
1.5.7. Risk, Safety, Security, and Reliability Analysis
Systems Engineers contribute to identifying and managing conditions that could prevent a system from achieving required performance, safety, security, reliability, availability, or other critical qualities. Human skills are important for recognizing context-dependent hazards, interpreting technical consequences, evaluating acceptable levels of risk, understanding interactions among system elements, and making consequential engineering judgments. Agentic skills can process requirements, architectures, failure information, risk records, standards, test results, and historical evidence to identify candidate risks, classify failure conditions, connect hazards with affected elements, compare mitigation alternatives, and monitor changes that may affect previously identified risks. Because risk evolves with the system, analysis can become a continuous human–agentic activity rather than a one-time assessment.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Identify consequential risks, interpret system-level effects, evaluate acceptability, select appropriate mitigations, and validate safety, security, or reliability judgments. | Analyze technical artifacts and evidence, identify candidate risks or failure conditions, classify them, connect them with system elements, and compare mitigation alternatives. | AI can systematically surface potential risks; human judgment determines their significance and acceptable treatment; design changes can trigger renewed agentic risk analysis. |
1.5.8. System Integration
System integration combines implemented system elements and verifies that their interfaces and interactions operate together as intended. Human skills are important for planning integration sequences, understanding dependencies, interpreting unexpected behavior, coordinating engineering teams, and determining whether integration results indicate local defects or broader system-level problems. Agentic skills can analyze integration plans, requirements, interface definitions, configuration information, logs, and test results; identify dependencies; organize integration evidence; detect anomalies; and suggest areas requiring further investigation. Integration results can therefore generate new information that moves repeatedly between automated analysis and human engineering judgment.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Plan integration strategy, coordinate disciplines, interpret unexpected interactions, diagnose system-level implications, and determine appropriate corrective actions. | Analyze integration plans and results, organize dependencies, process logs and evidence, detect anomalies, and identify possible areas for investigation. | Integration produces evidence for agentic analysis; AI findings can focus human investigation; engineering decisions can then modify configurations or integration plans for subsequent cycles. |
1.5.9. Verification and Validation
Verification and validation determine whether system requirements have been satisfied and whether the resulting system fulfills its intended needs and purposes under relevant operating conditions. Human skills are important for defining meaningful verification and validation strategies, determining acceptable evidence, interpreting results, understanding operational significance, and deciding whether unresolved deviations are acceptable. Agentic skills can map requirements to verification methods, generate preliminary procedures and test cases, organize evidence, compare observed results with expected outcomes, identify missing coverage, classify deviations, and maintain structured records of verification and validation status.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Define meaningful V&V strategies, determine acceptable evidence, interpret results, evaluate operational significance, and decide whether requirements and stakeholder needs have been satisfied. | Map requirements to V&V activities, generate preliminary procedures, organize evidence, compare outcomes with expectations, identify missing coverage, and classify deviations. | AI can prepare and analyze V&V information; human interpretation determines the significance of results; failures or uncertainties can trigger additional agentic analysis and revised verification activities. |
1.5.10. Requirements Traceability
Requirements traceability maintains relationships among stakeholder needs, system requirements, derived requirements, architecture elements, design decisions, risks, interfaces, implementation artifacts, verification activities, and validation evidence. Human skills are important for determining whether relationships are meaningful, resolving uncertain mappings, understanding the significance of changes, and validating that traceability represents the actual engineering logic of the system. Agentic skills can maintain large traceability structures, identify missing links, compare related artifacts, detect orphan requirements, analyze change propagation, and flag potentially affected elements when requirements or designs change.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Validate traceability logic, resolve uncertain relationships, interpret change implications, and ensure traceability represents actual engineering intent. | Maintain traceability relationships, identify missing or inconsistent links, detect orphan artifacts, compare changes, and identify potentially affected system elements. | AI can continuously maintain and inspect traceability; human changes to requirements or designs can trigger automated impact analysis; professional review validates resulting relationships. |
1.5.11. Technical Documentation
Systems Engineering generates and maintains requirements specifications, architecture descriptions, interface definitions, analysis reports, technical plans, risk records, verification and validation documentation, configuration information, design decisions, and other engineering artifacts. Human skills remain important for determining what information must be documented, interpreting technical meaning, validating consequential statements, and ensuring documentation accurately represents the system and engineering decisions. Agentic skills can consolidate information, generate preliminary documentation, normalize terminology, identify missing sections, compare related artifacts, detect inconsistencies, maintain templates, and update documentation when authorized source information changes.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Determine documentation requirements, communicate technical meaning, validate consequential statements, and ensure artifacts accurately represent engineering intent. | Consolidate technical information, draft and structure documents, normalize terminology, identify omissions, compare artifacts, and check consistency. | AI can perform substantial documentation processing; human review validates engineering meaning; corrections and new decisions can trigger regeneration and further consistency checking. |
1.5.12. Configuration and Change Management
Configuration and change management maintains controlled knowledge of system baselines, requirements, architectures, interfaces, designs, software, hardware, documentation, and other technical artifacts as they evolve. Human skills are important for evaluating why a change is necessary, understanding its system-level consequences, balancing competing priorities, and participating in decisions about whether changes should be accepted. Agentic skills can compare artifact versions, identify affected requirements or interfaces, trace dependencies, summarize proposed changes, detect inconsistencies, and generate preliminary impact analyses. Because a seemingly local modification can propagate through multiple system elements, agentic analysis can help broaden information coverage while professional judgment determines whether the change is technically and organizationally acceptable.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Evaluate proposed changes, understand system-level consequences, balance competing considerations, and participate in change decisions. | Compare versions, identify dependencies and affected artifacts, trace change propagation, summarize modifications, and generate preliminary impact analyses. | AI can identify potentially affected areas; human evaluation determines whether changes should proceed; approved modifications can trigger automated updates and renewed consistency analysis. |
1.5.13. Design Reviews and Technical Evaluation
Design reviews and technical evaluations examine requirements, architectures, designs, analyses, risks, interfaces, verification evidence, and other technical artifacts to determine whether system development is progressing appropriately. Human skills are central to challenging assumptions, interpreting engineering evidence, recognizing system-level weaknesses, asking context-sensitive questions, and determining whether technical maturity is sufficient for subsequent lifecycle activities. Agentic skills can prepare review materials, retrieve supporting evidence, compare artifacts against defined criteria, identify inconsistencies, detect unresolved items, summarize technical status, and generate candidate review questions.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Challenge assumptions, evaluate technical maturity, interpret evidence, identify consequential concerns, and determine whether the system is ready to progress. | Prepare review information, compare artifacts against criteria, identify inconsistencies or unresolved issues, summarize status, and generate possible review questions. | AI can perform extensive pre-review analysis; human reviewers interpret and challenge findings; review decisions can create new actions for subsequent agentic tracking and analysis. |
1.5.14. Technical Problem Investigation and Resolution
Technical problem investigation and resolution addresses failures, anomalies, requirement deviations, integration problems, unexpected system behavior, and other conditions requiring engineering diagnosis and corrective action. Human skills are important for framing the problem, interpreting incomplete evidence, developing hypotheses, understanding interactions among system elements, and determining the practical significance of possible causes. Agentic skills can process logs, test results, requirements, architecture information, change histories, previous incidents, and other evidence to identify patterns, correlate events, generate candidate hypotheses, retrieve related information, and organize possible corrective actions. Investigation can proceed iteratively as hypotheses are tested and new evidence becomes available.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Frame technical problems, interpret incomplete evidence, develop and evaluate hypotheses, understand system interactions, and select corrective actions. | Process technical evidence, identify patterns and correlations, retrieve related information, generate candidate causes, and organize possible corrective actions. | Human hypotheses can guide agentic evidence analysis; AI findings can redirect investigation; new tests or engineering observations can then become inputs for further analysis. |
1.5.15. Cross-Functional Engineering Collaboration
Systems Engineers collaborate with software, hardware, mechanical, electrical, safety, cybersecurity, quality, manufacturing, operations, project management, and other technical or organizational functions to maintain coherence across the complete system. Human skills are central to negotiating technical responsibilities, resolving disagreements, communicating system-level context, coordinating decisions, and balancing discipline-specific objectives against overall system needs. Agentic skills can summarize meetings, consolidate authorized engineering information, identify unresolved questions, organize action items, connect related information across technical artifacts, prepare internal briefings, and maintain accessible cross-disciplinary context.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Coordinate engineering disciplines, resolve disagreements, negotiate responsibilities, communicate system context, and participate in cross-functional decisions. | Summarize discussions, consolidate technical information, identify unresolved issues, organize actions, and maintain structured cross-disciplinary context. | AI-maintained context can support interdisciplinary discussions; engineering decisions and new information can subsequently update the context available for further agentic analysis. |
1.5.16. System Transition, Operation, and Lifecycle Support
Systems Engineering extends beyond development to system transition, deployment, operation, maintenance, support, modification, and eventual retirement. Human skills are important for understanding operational environments, interpreting user and operator needs, evaluating transition readiness, responding to unexpected field conditions, and determining how lifecycle decisions affect system effectiveness. Agentic skills can consolidate transition requirements, analyze operational records, summarize field issues, monitor recurring failures, retrieve maintenance information, compare observed performance with expected behavior, and organize lifecycle knowledge. Operational evidence can also reveal new requirements, risks, or design limitations that become inputs for subsequent engineering activity.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Evaluate transition readiness, understand operational context, interpret field conditions, coordinate lifecycle decisions, and determine the significance of operational feedback. | Organize transition information, process operational records, identify recurring issues, compare actual and expected performance, and maintain lifecycle knowledge. | AI can continuously process operational evidence; human interpretation determines its significance; field findings can trigger new requirements, analyses, or engineering changes. |
1.5.17. Technical Planning, Monitoring, and Control
Technical planning, monitoring, and control organizes Systems Engineering activities across the system lifecycle and evaluates whether technical work is progressing consistently with objectives, dependencies, constraints, risks, resources, and engineering milestones. Human skills contribute planning judgment, prioritization, organizational understanding, technical leadership, and the ability to respond to changes that cannot be anticipated completely in advance. Agentic skills can organize technical activities, identify dependencies, consolidate status information, track engineering actions, monitor requirements or verification progress, detect schedule or technical inconsistencies, and generate structured status summaries. As technical conditions change, plans can be repeatedly reassessed through human–agentic interaction.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Establish technical priorities, exercise engineering leadership, interpret project conditions, coordinate dependencies, and make consequential planning decisions. | Organize technical activities, track dependencies and status, consolidate engineering information, detect inconsistencies, and generate progress summaries. | AI can maintain detailed technical status; human evaluation determines necessary changes; revised priorities and plans can update subsequent agentic monitoring. |
1.5.18. Systems Engineering Process Improvement
Systems Engineering process improvement examines how requirements, architecture, modeling, reviews, documentation, integration, V&V, change management, and other engineering activities can be performed more effectively and consistently. Human skills contribute experiential knowledge, organizational understanding, judgment about practical feasibility, and the ability to determine whether a process change improves engineering outcomes rather than merely increasing procedural complexity. Agentic skills can analyze recurring documentation problems, workflow information, review findings, engineering defects, delays, change histories, and other evidence to identify patterns, compare current practices with defined procedures, detect recurring gaps, and generate candidate improvements.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Interpret process weaknesses, evaluate organizational feasibility, determine improvement priorities, and validate whether changes improve engineering effectiveness. | Analyze workflow and engineering evidence, identify recurring problems, compare practices with procedures, detect gaps, and generate candidate improvements. | AI can identify patterns across large collections of engineering activity; human judgment determines which improvements are practical; implemented changes can generate new evidence for continued agentic analysis. |
1.5.19. Systems, Domain, and Technology Knowledge
Systems Engineers need to maintain current knowledge of systems principles, engineering methods, relevant technologies, modeling approaches, standards, technical domains, lifecycle practices, and emerging developments that may affect system requirements, architectures, risks, integration, or operation. Human skills contribute professional understanding, experiential knowledge, interpretation of relevance, and the ability to connect new information with the specific characteristics of an engineered system. Agentic skills can retrieve and summarize technical literature and documentation, organize standards and technology updates, compare new information with existing engineering knowledge, explain unfamiliar concepts, filter information according to current priorities, and surface developments relevant to particular systems or technical decisions. As AI increasingly becomes both an engineering tool and a component of engineered systems, Systems Engineers may also need to understand how model capabilities, limitations, uncertainty, and agentic behavior interact with broader system requirements and lifecycle considerations.
| Human Skills | Agentic Skills | Dynamic Human–Agentic Interaction |
|---|---|---|
| Develop systems-level professional understanding, determine technical relevance, connect knowledge to particular systems, and apply experience to engineering decisions. | Retrieve, filter, summarize, compare, and organize systems, domain, standards, and technology information. | AI can continuously surface relevant technical knowledge; professional evaluation determines its significance; emerging system needs can guide subsequent agentic exploration. |
1.6. Human–Agentic Involvement Across Systems Engineering
The relative involvement of human and agentic skills differs not only between responsibilities but also among the tasks and subtasks contained within the same responsibility. The following matrix therefore represents common tendencies rather than fixed assignments.
| Responsibility | Typical Human Involvement | Typical Agentic Involvement | Common Interaction Pattern |
|---|---|---|---|
| Stakeholder Needs and System Requirements | Very High | High | Human-led discovery with agentic analysis |
| Requirements Analysis and Management | High | High | Agentic analysis with professional validation |
| System Architecture and Design | Very High | High | Human-led architecture with agentic exploration |
| System Modeling and Analysis | High | High | Iterative modeling and engineering interpretation |
| Interface Definition and Management | High | High | Agentic information management with human coordination |
| System and Decision Analysis | High | High | Agentic comparison with professional judgment |
| Risk, Safety, Security, and Reliability Analysis | Very High | High | Agentic analysis with consequential human judgment |
| System Integration | Very High | High | Human-led integration with agentic analysis |
| Verification and Validation | High | High | Collaborative evidence generation and evaluation |
| Requirements Traceability | Medium–High | Very High | Agentic-heavy traceability with human validation |
| Technical Documentation | Medium–High | Very High | Agentic-heavy documentation with professional review |
| Configuration and Change Management | High | High | Agentic impact analysis with human decision-making |
| Design Reviews and Technical Evaluation | Very High | High | Agentic preparation with human-led evaluation |
| Technical Problem Investigation and Resolution | Very High | High | Iterative human–agentic investigation |
| Cross-Functional Engineering Collaboration | Very High | Medium–High | Human coordination supported by persistent AI context |
| System Transition, Operation, and Lifecycle Support | High | High | Continuous operational analysis and human interpretation |
| Technical Planning, Monitoring, and Control | High | High | Agentic monitoring with human technical leadership |
| Systems Engineering Process Improvement | High | High | Agentic pattern analysis with human process judgment |
| Systems, Domain, and Technology Knowledge | High | High | Continuous agentic information processing and human learning |
1.7. Why the Division of Work Changes
The boundary between human and agentic participation should not be treated as permanent. It can change from one system, lifecycle stage, engineering responsibility, task, subtask, or moment of execution to another.
| Allocation Factor | Effect on Human–Agentic Work |
|---|---|
| Task complexity | Complex system-level work may require decomposition, repeated reasoning, multidisciplinary analysis, specialist involvement, or greater human supervision. |
| Task consequence | Decisions involving significant safety, security, technical, financial, operational, regulatory, or organizational consequences may require stronger human validation or authority. |
| Available information | Agentic performance depends on access to sufficiently complete, current, relevant, consistent, and authorized engineering information. |
| Model capability | More capable models may perform analytical or generative subtasks that previously required greater human effort, while weaker models may require closer supervision. |
| Agentic skill availability | The existence of appropriate reusable or dynamically constructed engineering skills affects what can be delegated effectively. |
| Tool access | Access to requirements repositories, modeling environments, simulation tools, configuration systems, engineering databases, code execution, test systems, search, or other tools can substantially change agentic capability. |
| Output reliability | Weak, inconsistent, unsupported, technically incorrect, or uncertain outputs can cause work to be returned to the Systems Engineer or another engineering specialist. |
| Intermediate results | New analyses, tests, integration results, failures, or design findings can change objectives, reveal constraints, create additional subtasks, or alter the appropriate division of work. |
| System criticality | Safety-critical, security-critical, mission-critical, or otherwise consequential systems may require greater human oversight, independent verification, or controlled engineering processes. |
| Lifecycle context | The appropriate division of work can differ substantially among concept development, design, implementation, integration, verification, operation, maintenance, and retirement. |
| Organizational policy | Governance, security, confidentiality, quality, compliance, configuration-control, and engineering approval rules may restrict particular forms of delegation. |
| Authority and accountability | Technical capability does not automatically provide organizational or professional authority to approve requirements, accept risks, authorize changes, certify systems, or make consequential engineering decisions. |
These factors can operate simultaneously. A task that is largely agentic under one set of conditions may become human-led under another, while improvements in models, agentic skills, engineering tools, information systems, and organizational processes may gradually change the practical boundary over time.
1.8. Developing Systems Engineering Capability with Agentic AI
The increasing availability of agentic skills does not eliminate the need for traditional Systems Engineering capabilities. Systems thinking, engineering knowledge, requirements reasoning, architecture development, modeling, technical judgment, interdisciplinary communication, risk understanding, lifecycle reasoning, and professional accountability remain important because they help the Systems Engineer understand what should be accomplished, determine whether generated results are technically appropriate, and connect individual engineering activities with the behavior and objectives of the complete system.
At the same time, professional capability increasingly includes the ability to work effectively with agentic systems. A Systems Engineer may need to determine which tasks can be delegated, provide the necessary system context and constraints, select appropriate agentic skills, evaluate intermediate and final results, identify unsupported assumptions or technically inconsistent outputs, request refinement, combine outputs from multiple agentic activities, recognize when specialist or human intervention is necessary, and incorporate validated results into subsequent engineering work. Competence therefore increasingly includes not only performing an engineering task directly but also orchestrating how that task should be performed across available human and agentic capabilities.
This creates a broader model of professional capability:
Professional Capability = Human Skills + Agentic Skills + Human–Agentic Coordination
The proportions of these components will differ across responsibilities and will continue to change as models, engineering technologies, systems, organizations, and professional practices evolve.
1.9. The Evolving Boundary Between Human and Agentic Skills
The relationship between a Systems Engineer and Agentic AI is best understood as an adaptive division of work rather than a fixed separation between human tasks and AI tasks. Some subtasks may be performed largely through agentic skills, especially where large amounts of requirements, technical documentation, model information, interface definitions, traceability relationships, test evidence, configuration information, or engineering knowledge must be retrieved, filtered, classified, summarized, compared, generated, checked, or repeatedly updated. Other activities may require substantial human participation because they involve systems thinking, incomplete or conflicting evidence, interdisciplinary negotiation, technical accountability, safety or security consequences, architectural judgment, acceptance of risk, or consequential engineering decisions. Between these areas lies a much larger collection of tasks in which human and agentic skills can interact through delegation, information exchange, analysis, review, correction, evaluation, refinement, escalation, and repeated reassignment.
The resulting work arrangement is therefore better represented as an adaptive human–agentic workflow than as a simple sequence in which either the professional or AI performs an entire responsibility independently. At one stage of system development, an AI agent may perform much of the requirements processing, traceability analysis, documentation checking, or technical information retrieval while the Systems Engineer supervises and validates the result. At another stage, the Systems Engineer may lead architectural decisions, design reviews, integration activities, or system-level risk evaluation while AI performs only supporting analysis, retrieval, documentation, or evidence organization. A subtask can move from human execution to agentic execution, return to human review, and later be delegated again as requirements change, system models evolve, integration results emerge, new risks are identified, operational evidence becomes available, or different engineering skills become necessary.
The long-term value of Agentic AI in Systems Engineering therefore lies not only in automating individual engineering activities. It lies in expanding the set of usable capabilities available to the professional and enabling those capabilities to be dynamically selected, combined, evaluated, and reallocated according to the evolving requirements of the system and its lifecycle. Systems Engineers who understand both the enduring value of professional human skills and the rapidly developing possibilities of agentic skills can use this changing division of work to improve information coverage, traceability, analytical efficiency, consistency, adaptability, and systems-level engineering effectiveness without reducing the occupation to either traditional human execution or fully automated AI performance.