Agent architecture and orchestration
I design multi-agent systems that do real work, not demos.
Supervisors, orchestrator-worker patterns, tool use, MCP and human-in-the-loop review, built on LangGraph, the Claude Agent SDK, Google ADK and Copilot agents. I decide what an agent should do, what deterministic code should do instead, and where a person signs off.
- Case studies
- 7
- Log entries
- 18
- Domains
- Telecom, Banking and financial services, Quality engineering, Legacy modernization, Conversational and voice AI, Developer platforms and tooling
- Core skills
- Multi-agent orchestration, Claude Agent SDK, LangGraph, Model Context Protocol, Realtime voice agents, Copilot custom agents and skills
Case studies
Multi-agent legacy documentation
An orchestrator, a deterministic analyzer, parallel explorer agents and an independent reviewer turn an undocumented codebase into a verified, 16-section technical report.
Client demo, embedded systems, Aug 2026 – Oct 2026
Read the case studyAgentic legacy modernization and code intelligence
Multi-agent workflows that moved a Visual Basic and Crystal Reports estate to Java 21 and JasperReports, guided by a GraphRAG code-intelligence platform.
Enterprise bank, 2025 – 2026
Read the case studyEstimation platform for an enterprise bank
A five-agent supervisor turns unstructured requirements and intake tickets into effort-scored plans, with SME review and an immutable audit trail.
Enterprise bank, 2025
Read the case studyRealtime voice agent for banking onboarding
A speech-to-speech agent that walks new customers through account opening, and hands off to a person when it isn't sure.
Enterprise bank, proof of concept, Dec 2024
Read the case studyAutonomous testing for a conversational voice agent
An orchestrator and parallel worker agents hold multi-turn conversations with a production voice assistant, while a design agent turns call-flow diagrams into test cases.
National telecom carrier, Jun 2026 – Jul 2026
Read the case studyCustom Copilot agents for test-automation teams
Custom agents and skills that generate UI, API and contract tests in each team's own conventions, rolled out with hands-on sessions and a searchable catalog of 27 agents.
Enterprise test-automation teams, Mar 2026 – Sep 2026
Read the case studyAgent and skill distribution platform
Turned copy-pasted AI agent files into versioned packages from a single registry, with a cross-platform installer and a staleness report.
National telecom carrier, May 2026 – present
Read the case studyEvidence from the log
- 7 Oct 2026
Demoed multi-agent documentation for legacy firmware
An orchestrator splits the codebase across explorer agents, and an independent reviewer agent rejects any claim it can't trace to a cited line.
AI angleShows an agent system whose output an engineer can trust, because a second agent verifies every claim against the source before it reaches the report.
- 22 Sep 2026
Launched a searchable catalog of 27 agents
Mapped every agent and skill to one of seven SDLC phases and published them in a docs portal, so engineers find an agent before they write a new one.
AI angleTurns a pile of agents into a toolkit. Reuse rises, duplicates fall, and the programme can see which SDLC phases still have no agent support.
- 15 Sep 2026
Added a reviewer agent to the documentation pipeline
Early drafts read well but overstated what the code did. A separate reviewer agent now checks each section against the source and sends unsupported claims back.
AI angleGenerator-reviewer separation applied to hallucination. The same model is never trusted to check its own work.
- 25 Aug 2026
Moved agent distribution to a single registry
Copies of agent files had drifted across repositories. Now one registry generates every manifest, and packages are versioned and reviewed before release.
AI angleAgents are now shipped like software. A prompt or skill fix reaches every team, and nothing changes agent behavior without review.
- 18 Aug 2026
Started the legacy documentation agents
Modernization projects stall because nobody understands the old code. Began an agent pipeline that combines deterministic static analysis with parallel LLM explorers.
AI angleDeterministic tools handle what must be exact (call graphs, registers) and agents handle what needs reading and explanation. The result becomes context for the agents that do the migration.
- 13 Aug 2026
Changed the rollout plan for code-review skills
Found that code-review skills only load from a repository's default branch, so teams testing on feature branches saw nothing. Rewrote the rollout steps around that.
AI angleKnowing how the agent platform actually loads skills is the difference between a rollout that works and teams deciding AI review "doesn't do anything".
- 7 Jul 2026
Generated test cases from call-flow designs in under 10 minutes
A Claude tool-use agent reads a call-flow design and writes one test case per branch, checked against a 22-case gold standard.
AI angleAn agent does the slow design work a tester used to do by hand, and a gold standard keeps it honest.
- 15 Jun 2026
Architected autonomous testing for a production voice assistant
Manual testing meant calling the assistant again for every branch. Designed an orchestrator that dispatches parallel worker agents to hold multi-turn test conversations.
AI angleAgents testing an agent. Worker agents play the customer, so coverage scales with compute instead of tester hours.
- 20 Apr 2026
Rolled out custom Copilot agents to test-automation teams
Wrote agents and skills that generate Playwright tests in each team's own conventions, and ran hands-on sessions so teams actually used them.
AI angleGeneric assistants write generic tests. Encoding each team's frameworks and conventions into agents is what made the output mergeable, and the sessions are what made people use it.
- 23 Mar 2026
Generated contract and API tests with agents
Extended agent workflows beyond UI tests to Pact contract tests and API suites, so service changes are caught before end-to-end runs.
AI angleMoves AI test generation down the test pyramid, where tests are faster and cheaper, instead of only producing more slow UI tests.
- 9 Feb 2026
Caught 9 of 9 injected bugs with risk-based test design
Ran a standard test set and an agent-assisted, risk-based set against the same feature with seeded bugs. The standard set caught 1.
AI angleHard evidence that how agents are directed matters more than how many tests they write.
- 19 Jan 2026
Mapped a legacy codebase into a knowledge graph
Turned repositories into a Neo4j graph of modules, calls and data flows, so engineers and agents query structure instead of grepping. Code navigation got 70% faster.
AI angleGraphRAG over code. Agents get relationships (who calls what, where data flows) that plain vector search misses.
- 8 Dec 2025
Finished the report migration ahead of plan
Crystal Reports moved to JasperReports with multimodal agents that compare old and new report output. The modernization ran 70% faster than planned.
AI angleA vision-capable agent checks the migrated report against the original, which is what made automated migration safe enough to trust.
- 10 Nov 2025
Moved Visual Basic modules to Java 21 with spec-driven agents
Ad-hoc AI coding gave inconsistent results. Each module now gets a written spec first, then agents implement and test against it.
AI angleSpec-driven development makes agent output repeatable. The spec is the contract the agent is held to, and the CI loop lets it correct itself.
- 18 Aug 2025
Cut manual estimation effort by 80%
The estimation workflow went live with SME review and an immutable audit trail. Estimators now review and adjust a cited draft instead of starting from a blank sheet.
AI angleAgents do the research and drafting, and a person keeps the decision. Human-in-the-loop plus an audit trail is what let a bank put it into production.
- 26 May 2025
Designed a five-agent estimation supervisor
Estimates started from long, unstructured requirement documents and varied by estimator. Built a LangGraph supervisor coordinating research and estimation agents.
AI angleA supervisor pattern with typed state and rollback, so each agent has one job and the workflow can recover when one fails.
- 3 Mar 2025
Built an LLM crawler for 20+ bank websites
Selector-based scrapers broke every time a site changed. The crawler lets a model read rendered pages into a fixed schema instead.
AI angleUses the LLM for what it's good at, reading messy pages like a person, and constrains it with a fixed output schema.
- 9 Dec 2024
Built a realtime voice agent for banking onboarding
A LiveKit and OpenAI Realtime voice agent that walks new customers through account opening, with handoff to a human when it isn't sure.
AI angleSpeech-to-speech agents with a confidence-based handoff. Building one taught me where voice agents fail, which later shaped how I test them.