Product Management Capability Model & RASCI Matrix

Roles, Responsibilities & AI Agent Mapping Domain: Finance Broker Management & Financial Wealth Management Platform: Paperclip.ai Date: April 2026

1. Approach and methodology

1.1 Frameworks used

This capability model is synthesised from four established Product Management frameworks, adapted for a software development organisation operating in financial services:

ISPMA Software Product Management Framework (V.2.0) — The international standard for software product management, covering product strategy, product planning, strategic management, and orchestration across development, marketing, sales, and services.
Pragmatic Framework for Product Managers — An academically grounded framework mapping PM activities that impact efficiency, business growth, and user satisfaction.
280 Group Product Management Maturity Model (PMMM) — Evaluates organisational maturity across product strategy, planning, development, and lifecycle management.
Product Management Competency Framework (Mind the Product / Spotify model) — Covers competency areas for individual PMs and team-level capability assessment.

1.2 RASCI model definition

The RASCI model extends the standard RACI with a Supportive role:

Letter	Role	Definition
R	Responsible	Does the work. Executes the task and delivers the output.
A	Accountable	Owns the outcome. Has final decision authority and sign-off. Only one A per capability.
S	Supportive	Actively assists the Responsible party. Provides resources, tools, or partial work products.
C	Consulted	Provides input before or during the work. Two-way communication.
I	Informed	Kept updated on progress and outcomes. One-way communication.

1.3 Roles defined

Six roles are mapped across the capability model. Four are AI agents operating within Paperclip.ai; two are human roles. AI Agent Roles:

Role	Abbreviation	Scope
CPO Agent	CPO-A	Orchestrator agent. Decomposes goals, delegates to specialist agents, enforces governance, produces roadmaps and OKRs.
Governance Agent	GOV-A	Standards enforcement agent. Monitors PM process adherence, manages audit trails, enforces quality gates and templates.
Operations Agent	OPS-A	Product operations agent. Manages analytics, feedback synthesis, documentation, release coordination, and experiment infrastructure.
Research Agent	RES-A	Discovery and intelligence agent. Conducts competitive analysis, market scanning, customer research synthesis, and regulatory monitoring.

Human Roles:

Role	Abbreviation	Scope
Human Product Leader	HPL	Senior human with strategy veto, compliance sign-off, and stakeholder relationship authority. Accountable for decisions with regulatory, strategic, or relationship consequences.
Product Specialist (Panel)	PS	Domain experts in broker management and wealth management. Execute discovery, delivery, and collaboration work. Operate as a panel of 2-3 for high-stakes decisions.

2. Product Management Capability Model

The model comprises 7 domains and 28 capabilities. Each capability represents a distinct function that must be covered for a mature PM practice.

Domain 1: Product Strategy

Defines the long-term direction, market positioning, and business model for each product line.

#	Capability	Description	Key Outputs
1.1	Product vision & mission	Defining the long-term purpose and direction of each product line (broker management, wealth management)	Vision statement, product charter, strategic narrative
1.2	Market & competitive analysis	Ongoing assessment of competitor products, fintech market trends, and positioning	Competitive landscape reports, market maps, positioning briefs
1.3	Business model & pricing	Revenue model design, pricing strategy, and financial viability assessment	Business cases, pricing models, revenue forecasts
1.4	Portfolio management	Balancing investment across broker management and wealth management product lines	Portfolio scorecards, investment allocation recommendations

Domain 2: Product Planning

Translates strategy into executable plans with measurable goals.

#	Capability	Description	Key Outputs
2.1	Roadmap management	Creating, maintaining, and communicating product roadmaps	Product roadmaps (now/next/later), theme-based roadmaps
2.2	Release planning	Scoping releases, defining timelines, coordinating dependencies	Release plans, dependency maps, scope documents
2.3	OKR & goal setting	Defining measurable objectives and key results per product	OKR frameworks, goal tracking dashboards
2.4	Capacity & resource planning	Aligning team capacity with roadmap commitments	Capacity plans, resource allocation matrices

Domain 3: Product Discovery

Understanding customer problems and validating solutions before committing to delivery.

#	Capability	Description	Key Outputs
3.1	Customer research & insights	Broker/advisor interviews, surveys, usage analytics interpretation	Research reports, insight summaries, persona updates
3.2	Problem validation	Confirming that identified problems are real and worth solving	Problem statements, opportunity assessments
3.3	Solution ideation & prototyping	Generating and testing solution concepts with stakeholders	Wireframes, prototypes, concept test results
3.4	Experimentation & A/B testing	Designing experiments to validate hypotheses with data	Experiment designs, test results, statistical analyses

Domain 4: Product Delivery

Translating validated solutions into shipped product through engineering collaboration.

#	Capability	Description	Key Outputs
4.1	Requirements & specification	Writing PRDs, user stories, acceptance criteria	PRDs, user stories, acceptance criteria documents
4.2	Backlog management	Prioritising, grooming, and maintaining the product backlog	Prioritised backlogs, grooming session outputs
4.3	Sprint collaboration	Working with engineering during sprints, unblocking, clarifying	Sprint goals, daily standup participation, clarification logs
4.4	Quality assurance oversight	Ensuring delivered features meet specification and quality bars	QA sign-off checklists, defect triage decisions

Domain 5: Orchestration & Collaboration

Coordinating across functions and external partners to align around product goals.

#	Capability	Description	Key Outputs
5.1	Stakeholder management	Managing expectations of executives, sales, brokers, advisors	Stakeholder maps, communication plans, status updates
5.2	Cross-functional alignment	Coordinating with engineering, design, compliance, marketing	Alignment meeting outputs, decision logs, DACI records
5.3	Go-to-market coordination	Launch planning, enablement, partner communication	Launch plans, enablement materials, partner communications
5.4	Vendor & partner management	Managing third-party integrations and partner relationships	Partner scorecards, integration roadmaps, contract reviews

Domain 6: Product Operations

Operational capabilities that enable data-driven product management at scale.

#	Capability	Description	Key Outputs
6.1	Product analytics & metrics	Defining, collecting, and reporting on product KPIs	Dashboards, metric reports, anomaly alerts
6.2	Customer feedback management	Collecting, categorising, and synthesising feedback at scale	Feedback taxonomies, trend reports, prioritised opportunity backlogs
6.3	Competitive intelligence	Monitoring competitor moves and market shifts systematically	Weekly intelligence briefs, threat/opportunity assessments
6.4	Documentation & knowledge management	Maintaining product documentation, wikis, decision logs	Product wikis, decision logs, onboarding materials

Domain 7: Governance & Compliance

Ensuring the PM function itself operates to defined standards and meets regulatory requirements.

#	Capability	Description	Key Outputs
7.1	PM standards & process enforcement	Defining and enforcing PM best practices across teams	PM playbook, process templates, compliance reports
7.2	Regulatory compliance oversight	Ensuring products meet financial services regulations (FCA, ASIC, MiFID II)	Compliance checklists, regulatory impact assessments
7.3	Risk management	Identifying and mitigating product and delivery risks	Risk registers, mitigation plans, escalation protocols
7.4	Audit & traceability	Maintaining decision audit trails for regulatory and internal review	Audit logs, decision trails, change records

3. RASCI Matrix

3.1 Full RASCI assignment

Legend:  R = Responsible  A = Accountable  S = Supportive  C = Consulted  I = Informed
Roles:  CPO-A = CPO Agent | GOV-A = Governance Agent | OPS-A = Operations Agent
        RES-A = Research Agent | HPL = Human Product Leader | PS = Product Specialist

Domain 1: Product Strategy

#	Capability	CPO-A	GOV-A	OPS-A	RES-A	HPL	PS
1.1	Product vision & mission	S	I	I	C	A	R
1.2	Market & competitive analysis	A	I	S	R	C	C
1.3	Business model & pricing	C	I	S	S	A	R
1.4	Portfolio management	S	I	C	C	A	R

Pattern: Strategy is human-led. The Human Product Leader holds Accountability for vision, pricing, and portfolio. AI agents support through data gathering and analysis (Research Agent leads competitive analysis), but strategic choices remain human.

Domain 2: Product Planning

#	Capability	CPO-A	GOV-A	OPS-A	RES-A	HPL	PS
2.1	Roadmap management	R	S	S	C	A	C
2.2	Release planning	S	C	R	I	I	A
2.3	OKR & goal setting	R	S	S	C	A	C
2.4	Capacity & resource planning	S	I	R	I	A	C

Pattern: Planning is a collaboration zone. The CPO Agent generates roadmap drafts and OKR proposals. The Operations Agent handles release mechanics and capacity modelling. Humans retain Accountability for approving plans.

Domain 3: Product Discovery

#	Capability	CPO-A	GOV-A	OPS-A	RES-A	HPL	PS
3.1	Customer research & insights	I	I	S	R	I	A
3.2	Problem validation	I	I	S	S	C	A
3.3	Solution ideation & prototyping	I	I	I	S	C	A
3.4	Experimentation & A/B testing	I	I	R	S	I	A

Pattern: Discovery is human-anchored. Product Specialists hold Accountability for all discovery capabilities because they possess domain expertise in broker and wealth management workflows. The Research Agent synthesises data, and the Operations Agent manages experiment infrastructure, but human judgment validates problems and solutions.

Domain 4: Product Delivery

#	Capability	CPO-A	GOV-A	OPS-A	RES-A	HPL	PS
4.1	Requirements & specification	S	C	I	I	I	A
4.2	Backlog management	S	R	S	I	I	A
4.3	Sprint collaboration	I	I	I	I	I	A
4.4	Quality assurance oversight	I	R	S	I	I	A

Pattern: Delivery is Product Specialist-led with agent support. The Governance Agent is Responsible for backlog hygiene and QA enforcement (checking stories against standards, flagging incomplete acceptance criteria). Sprint collaboration is almost entirely human — daily standups and engineer unblocking cannot be delegated.

Domain 5: Orchestration & Collaboration

#	Capability	CPO-A	GOV-A	OPS-A	RES-A	HPL	PS
5.1	Stakeholder management	S	I	I	I	A	R
5.2	Cross-functional alignment	S	I	I	I	A	R
5.3	Go-to-market coordination	S	I	R	S	A	C
5.4	Vendor & partner management	I	I	I	C	A	R

Pattern: Orchestration is the most human-intensive domain. Stakeholder management, cross-functional alignment, and vendor relationships are irreducibly human. Agents prepare materials and coordinate logistics (CPO Agent drafts stakeholder updates, Operations Agent manages GTM checklists), but relationship-building and negotiation cannot be automated.

Domain 6: Product Operations

#	Capability	CPO-A	GOV-A	OPS-A	RES-A	HPL	PS
6.1	Product analytics & metrics	S	I	R	S	I	A
6.2	Customer feedback management	I	I	R	S	I	A
6.3	Competitive intelligence	A	I	S	R	I	C
6.4	Documentation & knowledge mgmt	S	R	S	I	I	C

Pattern: Operations is the most agent-intensive domain. Agents handle 80-90% of the operational workload — aggregating metrics, synthesising feedback, scanning competitors, and maintaining documentation. Product Specialists retain Accountability (reviewing and approving agent outputs) but do not perform the bulk of the work.

Domain 7: Governance & Compliance

#	Capability	CPO-A	GOV-A	OPS-A	RES-A	HPL	PS
7.1	PM standards & process enforcement	A	R	S	I	C	I
7.2	Regulatory compliance oversight	S	S	I	C	A	R
7.3	Risk management	S	R	S	C	A	C
7.4	Audit & traceability	S	R	S	I	A	I

Pattern: Governance is split. Process standards and audit trails are highly automated (Governance Agent enforces templates, maintains logs, checks adherence). Regulatory compliance is human-critical — the Human Product Leader is Accountable with Product Specialists as the Responsible domain experts. In financial services, no agent should autonomously determine compliance.

4. Pitcher / Catcher Diagram

The Pitcher/Catcher model shows the collaboration dynamic for each capability. The Pitcher initiates and delivers the primary work product. The Catcher receives, validates, and acts on it. This model makes handoff points explicit and ensures no capability falls between roles.

4.1 How to read the diagram

AI AGENT ◄──────────────────┤ Handoff ├──────────────────► HUMAN
  (Pitcher)                     Point                    (Catcher)
                          ◄── or ──►
                          (Pitcher)                      (Catcher)

When AI pitches: The agent generates the work product (draft, analysis, report) and the human catches (reviews, validates, decides).
When Human pitches: The human initiates (interviews, negotiations, strategic choices) and the agent catches (records, structures, distributes).
When balanced (50/50): True collaboration — both parties contribute equally, with the handoff point in the middle.

4.2 Full Pitcher / Catcher mapping

                     AI AGENT                              HUMAN
                     pitches ◄───────────┼───────────► pitches
                          100%    50%    0%    50%    100%

DOMAIN 1: PRODUCT STRATEGY
1 Vision & mission         ██░░░░░░░░┼██████████████████  Human pitches (80%)
2 Market & competitive     █████████████████░░░░░░░░░░░░  Agent pitches (75%)
3 Business model & pricing ██████░░░░┼██████████████░░░░  Human pitches (70%)
4 Portfolio management     █████░░░░░┼███████████████░░░  Human pitches (75%)

DOMAIN 2: PRODUCT PLANNING
1 Roadmap management       ████████████░░░░┼████████░░░░  Agent pitches (60%)
2 Release planning         ███████████░░░░░┼█████████░░░  Agent pitches (55%)
3 OKR & goal setting       █████████████░░░┼███████░░░░░  Agent pitches (65%)
4 Capacity & resource      ██████████░░░░░░┼██████████░░  Balanced (50%)

DOMAIN 3: PRODUCT DISCOVERY
1 Customer research        █████████░░░░░░░┼███████████░  Human pitches (55%)
2 Problem validation       ██████░░░░┼██████████████░░░░  Human pitches (70%)
3 Solution ideation        █████░░░░░┼███████████████░░░  Human pitches (75%)
4 Experimentation          ███████████░░░░░┼█████████░░░  Agent pitches (55%)

DOMAIN 4: PRODUCT DELIVERY
1 Requirements & spec      ██████████████░░┼██████░░░░░░  Agent pitches (70%)
2 Backlog management       █████████████████┼█████░░░░░░  Agent pitches (75%)
3 Sprint collaboration     ███░░░░░░░┼█████████████████░  Human pitches (85%)
4 QA oversight             ████████████░░░░┼████████░░░░  Agent pitches (60%)

DOMAIN 5: ORCHESTRATION & COLLABORATION
1 Stakeholder management   ███░░░░░░░┼█████████████████░  Human pitches (85%)
2 Cross-functional align   ███░░░░░░░┼█████████████████░  Human pitches (85%)
3 Go-to-market coord       ██████████░░░░░░┼██████████░░  Balanced (50%)
4 Vendor & partner mgmt    ██░░░░░░░░┼██████████████████  Human pitches (90%)

DOMAIN 6: PRODUCT OPERATIONS
1 Product analytics        █████████████████░░┼███░░░░░░  Agent pitches (85%)
2 Customer feedback        ████████████████░░░┼████░░░░░  Agent pitches (80%)
3 Competitive intelligence █████████████████░░┼███░░░░░░  Agent pitches (85%)
4 Documentation            ██████████████████░░┼██░░░░░░  Agent pitches (90%)

DOMAIN 7: GOVERNANCE & COMPLIANCE
1 PM standards enforcement █████████████████░░┼███░░░░░░  Agent pitches (85%)
2 Regulatory compliance    ████░░░░░░┼████████████████░░  Human pitches (80%)
3 Risk management          █████████░░░░░░░┼███████████░  Human pitches (55%)
4 Audit & traceability     ██████████████████░░┼██░░░░░░  Agent pitches (90%)

4.3 Pitcher / Catcher summary

Category	Capabilities	Pattern
Agent pitches, Human catches	1.2, 2.1, 2.2, 2.3, 3.4, 4.1, 4.2, 4.4, 6.1, 6.2, 6.3, 6.4, 7.1, 7.4	14 capabilities (50%). Agent generates work product; human reviews and approves.
Human pitches, Agent catches	1.1, 1.3, 1.4, 3.1, 3.2, 3.3, 4.3, 5.1, 5.2, 5.4, 7.2, 7.3	12 capabilities (43%). Human initiates; agent records, structures, and distributes.
Balanced collaboration	2.4, 5.3	2 capabilities (7%). True 50/50 collaboration with iterative handoffs.

5. Gap Analysis and Role Design Recommendations

5.1 Critical gaps

GAP 1: No human Product Leader defined The RASCI reveals 12 capabilities where a human holds Accountability. Without a designated Human Product Leader (even part-time), there is no one to exercise veto authority on strategy, compliance, and portfolio decisions. This is the single most critical gap to address before deploying any AI agents. GAP 2: Regulatory compliance requires human sign-off Capability 7.2 (Regulatory compliance oversight) is scored at only 20% agent automation. In financial services, product decisions touching FCA, ASIC, or MiFID II rules require a named individual with compliance authority. An agent can flag risks and surface regulations, but cannot bear legal accountability.

5.2 Role design recommendations

REC 1: Product Specialist role evolution Product Specialists currently hold Accountability for 16 of 28 capabilities but have no governance framework. The RASCI shows they need to become agent supervisors: reviewing AI-generated PRDs, validating feedback synthesis, and approving backlog changes. New skills needed include prompt engineering, agent output evaluation, and AI-assisted decision-making. REC 2: New role — Agent Operations Lead With 4 AI agents (CPO, Governance, Operations, Research), someone must monitor agent health, manage API costs, tune prompts, and handle agent failures. This role does not exist today and must be created. It can initially be a secondary responsibility for one Product Specialist, growing to a dedicated role as agent usage scales. REC 3: Product Specialist panel structure For capabilities where the Human Product Leader is Accountable and Product Specialists are Responsible (1.1, 1.3, 1.4, 5.1, 5.2, 5.4, 7.2), a formal panel of 2–3 specialists should review and recommend. This creates collective domain judgment rather than single-point-of-failure decisions. The panel structure is especially important for regulatory compliance in financial services.

5.3 Coverage gaps

COV 1: Sprint collaboration is under-covered by agents Capability 4.3 (Sprint collaboration) is only 15% automatable. Daily standups, unblocking engineers, and real-time technical clarification require human presence. Ensure Product Specialists have sufficient capacity for this function — it cannot be delegated. COV 2: Stakeholder and vendor management are human-only Capabilities 5.1, 5.2, and 5.4 are 10–15% automatable. Agents can prepare briefing materials and model scenarios, but relationship management with brokers, wealth advisors, and technology partners is irreducibly human. These must be explicitly assigned to named individuals with allocated time. COV 3: Discovery capabilities need human anchoring Capabilities 3.2 (Problem validation) and 3.3 (Solution ideation) are 25–30% automatable. These require empathy, contextual understanding of broker workflows, and creative thinking. Agents augment with data synthesis but cannot lead discovery. Ensure dedicated discovery time for specialists — at least 20% of their week.

6. Role Definition Summary

6.1 AI Agent roles — Paperclip.ai configuration

Agent	Reports to	Budget tier	Approval gates	Primary domain coverage
CPO Agent (Orchestrator)	Human Product Leader	High	Strategy outputs, roadmap changes, OKR proposals	Domains 1, 2 (planning layer)
Governance Agent	CPO Agent	Medium	PM standard changes, compliance flag overrides	Domains 4, 7 (enforcement layer)
Operations Agent	CPO Agent	Medium	Dashboard publication, feedback report distribution	Domain 6 (operations layer)
Research Agent	CPO Agent	Medium	Competitive brief publication, regulatory alert distribution	Domains 1, 3 (intelligence layer)

6.2 Human roles — Job design

Human Product Leader (Part-time or fractional initially)

Accountable for 12 capabilities across strategy, planning, orchestration, and compliance
Veto authority on all agent-generated strategy recommendations
Compliance sign-off authority for financial services regulations
Stakeholder relationship ownership with executives, sales, and broker/advisor partners
Chairs the Product Specialist panel for high-stakes decisions

Product Specialist (2–3 individuals, forming a rotating panel)

Accountable for 16 capabilities across discovery, delivery, and orchestration
Domain expert in either broker management or wealth management (or both)
Primary interface with engineering teams during sprints
Validates all agent-generated outputs before external distribution
Participates in panel reviews for strategy, compliance, and partner decisions

Agent Operations Lead (Secondary role initially, growing to dedicated)

Monitors AI agent health, cost, and output quality
Tunes prompts and agent configurations based on output evaluation
Manages Paperclip.ai platform: budgets, approval gates, audit logs
Escalates agent failures or anomalous behaviour to Human Product Leader
Maintains the PM framework templates that agents enforce

7. Sources and references

Source	Year	Relevance
ISPMA Software Product Management Framework V.2.0	2024	Primary capability model reference. Covers product strategy, planning, orchestration, and cross-functional collaboration.
280 Group Product Management Maturity Model	2024	Maturity assessment across strategy, planning, development, and lifecycle management.
Pragmatic Framework for Product Managers (arXiv)	2025	Academically grounded PM activity framework for efficiency and business impact.
Mind the Product Competency Framework	2025	Skills-based PM assessment covering discovery, delivery, and leadership competencies.
GitLab Product Management CDF	2026	Open-source career development framework with detailed capability descriptions per PM level.
MIT Sloan / BCG — Emerging Agentic Enterprise	2025	Research on AI agent adoption, governance requirements, and organisational design.
McKinsey — The Agentic Organisation	2025	Predictions on agent task completion capability and governance framework requirements.
Deloitte — Agent Ops Prediction	2025	Forecast of dedicated agent operations teams and new role definitions.
AI PM Tools Directory — 51-Tool Analysis	2026	Benchmarking of agentic capabilities across PM tooling landscape.
Paperclip.ai Documentation	2026	Platform-specific governance features: approval gates, budget controls, goal alignment, audit trails.

This document defines the roles, responsibilities, and coverage model for the Product Management function. It should be reviewed quarterly as agent capabilities mature and organisational needs evolve. The RASCI assignments represent initial configuration — actual role boundaries should be adjusted based on measured agent output quality and the maturity of the PM governance framework.

Getting Started

Frameworks

Processes

Templates

Research

Tech Reports

​Product Management Capability Model & RASCI Matrix

​1. Approach and methodology

​1.1 Frameworks used

​1.2 RASCI model definition

​1.3 Roles defined

​2. Product Management Capability Model

​Domain 1: Product Strategy

​Domain 2: Product Planning

​Domain 3: Product Discovery

​Domain 4: Product Delivery

​Domain 5: Orchestration & Collaboration

​Domain 6: Product Operations

​Domain 7: Governance & Compliance

​3. RASCI Matrix

​3.1 Full RASCI assignment

​Domain 1: Product Strategy

​Domain 2: Product Planning

​Domain 3: Product Discovery

​Domain 4: Product Delivery

​Domain 5: Orchestration & Collaboration

​Domain 6: Product Operations

​Domain 7: Governance & Compliance

​4. Pitcher / Catcher Diagram

​4.1 How to read the diagram

​4.2 Full Pitcher / Catcher mapping

​4.3 Pitcher / Catcher summary

​5. Gap Analysis and Role Design Recommendations

​5.1 Critical gaps

​5.2 Role design recommendations

​5.3 Coverage gaps

​6. Role Definition Summary

​6.1 AI Agent roles — Paperclip.ai configuration

​6.2 Human roles — Job design

​7. Sources and references

Product Management Capability Model & RASCI Matrix

1. Approach and methodology

1.1 Frameworks used

1.2 RASCI model definition

1.3 Roles defined

2. Product Management Capability Model

Domain 1: Product Strategy

Domain 2: Product Planning

Domain 3: Product Discovery

Domain 4: Product Delivery

Domain 5: Orchestration & Collaboration

Domain 6: Product Operations

Domain 7: Governance & Compliance

3. RASCI Matrix

3.1 Full RASCI assignment

Domain 1: Product Strategy

Domain 2: Product Planning

Domain 3: Product Discovery

Domain 4: Product Delivery

Domain 5: Orchestration & Collaboration

Domain 6: Product Operations

Domain 7: Governance & Compliance

4. Pitcher / Catcher Diagram

4.1 How to read the diagram

4.2 Full Pitcher / Catcher mapping

4.3 Pitcher / Catcher summary

5. Gap Analysis and Role Design Recommendations

5.1 Critical gaps

5.2 Role design recommendations

5.3 Coverage gaps

6. Role Definition Summary

6.1 AI Agent roles — Paperclip.ai configuration

6.2 Human roles — Job design

7. Sources and references