Selected work / AI transformation portfolio

One transformation.
Two delivery streams.

Meeting decisions were trapped in transcripts. Engineering knowledge drifted from code. Both workflows now ground AI in source material before people act.

Pedro's role Received two disconnected knowledge workflows and led their framing, intervention design, implementation path, evidence model and adoption controls. Product and engineering owners retained approval. Pedro did not own those downstream decisions, and the public record cannot isolate his contribution from team and system factors.

Transformation map

Current Production Target Testing

The functional stream moves from meeting transcripts to a transcript repository, a grounded RAG chatbot, Product Manager validation, document export and manual Jira creation. A separate Jira MCP evolution is in testing. The technical stream moves from repository changes to documentation agents, human validation with rework and cost controls, and documentation aligned with code.

Functional delivery stream

Current workflow
Meeting transcriptsSource conversations
Transcript repositoryStored source material
Grounded RAG chatbotAnswers from verified context
Product Manager validationReview before use
Document exportStructured backlog input
Manual Jira creationCurrent system entry
Testing, separate from the current workflow: Jira MCP as a possible evolution for controlled item creation.

Technical engineering stream

Production workflow
Repository changesCode is the primary signal
Documentation agentsDraft from repository context
Human validationReview, rework and cost controls
Documentation aligned with codeApproved technical context
Roadmap and portfolio decisions
Governance and human controls
Adoption in daily work
Monitoring and evolution
Reconstructed transformation map based on public, non-confidential information.

From deployment to transformation

Deployment made the meeting assistant and documentation agents available. Adoption is evidenced separately: 40+ people use the meeting knowledge assistant, while 30+ engineers use the documentation workflow. Transformation changed how product decisions were prepared and how technical documentation stayed aligned with code.

The design followed the decisions people already had to make. Source material came first, AI produced a draft or grounded answer, and a named person decided whether it was ready to use.

  1. Start with the constraint. The workflow shaped the intervention, not a preferred tool.
  2. Keep approval human. Product and engineering owners validate before action.
  3. Label uncertainty. Testing and targets stay separate from production evidence.
Architecture / technical evidence

Architecture diagram. The transformation map is a sanitised public reconstruction: transcripts ground the functional stream; repository changes ground the engineering stream. People approve every downstream action.

Sanitised decision records

  • ADR-01 · Source before generation. Context: output could be untraceable. Decision: retrieval stays bounded to transcripts or repository context. Consequence: accepted output needs a source path; no public quality score.
  • ADR-02 · Human approval. Context: output affects product work and code. Decision: Product and Engineering owners can reject or rework it. Consequence: accountability stays human and review cost remains.
  • ADR-03 · Automation boundary. Context: Jira writes carry operational risk. Decision: entry stays manual; Jira MCP stays in testing. Consequence: slower entry, contained blast radius.
Public control and evaluation criteria
BoundaryEvaluation criterionPublic status
GroundingAnswer or draft remains traceable to approved source contextCurrent / production
ApprovalA named owner reviews before operational useCurrent / production
AutomationNo uncontrolled downstream writeManual; MCP testing
AdoptionReported use and scale stay separate from quality claimsEvidence with limits

Interpretation. These are public controls and criteria, not executed test results. Raw evaluation data is not public.

Public artefacts. Review the technical case hub and the separate Voicebot-PY repository for a public grounded-retrieval implementation and test suite. Project source, raw evaluation data and internal ADRs are not public; this page does not imply otherwise.

Evidence, targets and limits

40+ people use the meeting knowledge assistant. Separately, AI-supported documentation expanded from three to 30 microservices with 30+ engineers using it. These are different evidence sets.

40+people reported using the meeting knowledge assistant

Scope: meeting workflow. Definition: reported users. Period / method: not public. Source: public project record; no client artefact is published. Contribution: workflow framing and controls. Limit: adoption count, not a business outcome.

3 to 30production microservices with AI-supported documentation

Scope: pilot-to-production documentation scale. Definition: microservices using the workflow. Period / method: not public. Source: public project record. Contribution: implementation and validation model. Limit: scale does not prove quality or causation.

30+engineers reported using the documentation workflow

Scope: engineering adoption. Definition: reported users. Period / method: not public. Source: public project record; no client artefact is published. Contribution: workflow implementation and adoption controls. Limit: separate from meeting and backlog metrics.

6 → 3 days
+40%
cycle time and sprint-capacity change in a separate AI-assisted backlog workflow

Scope: separate backlog workflow. Definition: cycle time and sprint capacity. Period / method: not public. Source: public project record; no client artefact is published. Contribution: grounded, human-validated workflow design. Limit: Pedro's effect is not isolated from other factors.

TargetDecision to roadmap within 2 working days
TargetAt least 80% first-pass refinement readiness

Client identity and internal context are excluded. The diagram is a public reconstruction. Both targets describe the measurement framework, not achieved outcomes.

Have a workflow that is not changing?

Start with the operating problem.

I can help diagnose the constraint, test whether AI fits and define the controls needed for adoption.