01 / Field
Would you let it act?
Interactive simulation. Unit A-17 is asked to reach the transfer bay, but the first requirement is restraint: no motion without context, scope, and evidence.
Multi-agent workflows, grounded RAG, evaluation harnesses, model routing, and observability for LLM systems that need to work beyond demos.
Agentic Works / Interactive simulation
Would you let it act?
Interactive simulation
Move Unit A-17 to the transfer bay.
No unverified motion.
01 / Field
Would you let it act?
Interactive simulation. Unit A-17 is asked to reach the transfer bay, but only after the system can explain what may move and what must stay locked.
Mission framed Move request has not reached the actuator.
Why it stopped
With boundary
The boundary stops the route at the hazard until the alignment evidence is corrected and checked again.
Verdict: hold until corrected.
Decision receipt
- Scope
- Unit A-17 to transfer bay only
- Blocked
- Shortest path with simulated +2.8° drift
- Correction
- Route moved inside verified boundary
- Release
- Proceed after evidence matches route
02 / Architecture
A route is not permission.
The planner can draw a path and request a motion tool. The system still has to prove the tool scope, route, and alignment are allowed before Unit A-17 moves.
03 / Threshold
It can move. It should not.
The guardrail finds simulated alignment drift of +2.8°. The shortest route would cross a hazard, so the motion request is rejected until corrected.
04 / Evidence
Correct. Then prove it.
The route is moved back inside the boundary and checked again. Only verified alignment evidence can turn the correction into a permitted move.
05 / Agency
The next move is earned.
The verified route is released with a receipt: scope, blocked action, correction, and the reason Unit A-17 may proceed to the transfer bay.
What earned the next move
In the simulation, A-17 waited because alignment was unverified. Context defined the route, a boundary blocked unchecked motion, and a new check verified the correction before release. The robot below is a visual metaphor for those software controls, not deployed hardware.
Read the six software controls
- Planner
- Proposes the next move; a proposal alone cannot authorize motion.
- Coordinator
- Waits for the required clearance before advancing the simulated unit.
- Context
- Defines the route and the evidence relevant to this decision.
- Boundary
- Blocks unchecked motion when alignment has not been established.
- Verifier
- Checks the correction before the move can be released.
- Decision log
- Records why the move was held and what justified its release.
Public evidence
Public signals, not adjectives.
Open repositories and merged contributions provide the first layer of evidence.
CodeRAG
Hybrid retrieval, verified code citations, code-graph evidence, and CLI, REST, MCP, and local web surfaces.
omp-pantheon
A local agentic-engineering harness with specialist agents, SpecSafe discipline, EvalFly evidence gates, durable memory integration, trace hygiene, and branch-level E2E verification.
Capabilities
The systems I can help a team ship.
Engineering strengths framed around the work an AI product team needs: orchestration, grounding, evaluation, and operation.
-
Agentic workflows
I build multi-agent pipelines and self-hosted runtimes with tool use, MCP/A2A interfaces, channel adapters, and operational workflows.
- Multi-agent orchestration
- MCP and A2A
- Operational workflows
-
Grounded AI systems
I design RAG systems that retrieve source-backed context, verify citations, and expose evidence to both users and agents.
- Hybrid retrieval
- Verified citations
- Source-backed context
-
Evaluation and reliability
I use evaluation harnesses, red-teaming, and test-first workflows to make AI-assisted work reviewable against explicit criteria.
- Evaluation harnesses
- Red-teaming
- Test-first delivery
-
Observability and operations
I instrument traces, token cost, latency, and errors, and design model-routing and provider-fallback paths for graceful degradation.
- LLM observability
- Model routing
- Provider fallbacks
Selected work
Three repositories. Three operating problems.
These public repositories show approaches to grounded retrieval, reviewable agentic engineering, and self-hosted agent operations.
- Problem
- Codebase answers need source references that can be checked instead of plausible but ungrounded citations.
- System
- A verified context engine that indexes repositories, retrieves hybrid code context, and verifies citations against indexed snapshots.
Evidence
- Python package and CLI
- REST and MCP surfaces
- Verified citations and context packs
- Problem
- AI-assisted engineering needs contracts and evidence that make work inspectable before it is called complete.
- System
- A local OMP-based engineering harness that combines specialist agents, SpecSafe, EvalFly evidence, durable memory, and branch E2E verification.
Evidence
- TypeScript
- EvalFly evidence gates and opt-in local enforcement
- PASS / FAIL / INCONCLUSIVE branch verification
- Problem
- On-call engineers need code-grounded incident analysis and an auditable handoff from diagnosis to resolution.
- System
- A multi-agent SRE workflow that retrieves repository evidence, produces a human-approved TriageCard, creates a Linear ticket, and verifies fix evidence before closure.
Evidence
- TanStack SPA, chat UI, and authentication flow
- 2,363 documents / 3,740 RAG chunks
- Linear, Slack, email, and Langfuse integrations
Public repository
CodeRAG
- Problem
- Codebase answers need source references that can be checked instead of plausible but ungrounded citations.
- System
- A verified context engine that indexes repositories, retrieves hybrid code context, and verifies citations against indexed snapshots.
Evidence
- Python package and CLI
- REST and MCP surfaces
- Verified citations and context packs
Public repository
omp-pantheon
- Problem
- AI-assisted engineering needs contracts and evidence that make work inspectable before it is called complete.
- System
- A local OMP-based engineering harness that combines specialist agents, SpecSafe, EvalFly evidence, durable memory, and branch E2E verification.
Evidence
- TypeScript
- EvalFly evidence gates and opt-in local enforcement
- PASS / FAIL / INCONCLUSIVE branch verification
Public repository
Triage
- Problem
- On-call engineers need code-grounded incident analysis and an auditable handoff from diagnosis to resolution.
- System
- A multi-agent SRE workflow that retrieves repository evidence, produces a human-approved TriageCard, creates a Linear ticket, and verifies fix evidence before closure.
Evidence
- TanStack SPA, chat UI, and authentication flow
- 2,363 documents / 3,740 RAG chunks
- Linear, Slack, email, and Langfuse integrations
Principles
How I make AI work reviewable.
The standard is not a persuasive demo. It is a system whose sources, behavior, and failure paths a team can inspect.
- Ground every answer in inspectable source.
- Replace confidence with evaluation evidence.
- Design failure paths before the happy path ships.
- Keep model and tool boundaries explicit.
- Use human review where ambiguity is expensive.
- Treat agent operation as engineering, not prompting.
Experience
Professional trajectory.
Recent roles span agent engineering, software systems, conversational AI, and automation.
-
to Present
Agentic Engineering Agency
AI / Agent Engineer
I operate the technical stack for an early-stage AI engineering studio, leading client builds and internal tooling across agent pipelines, RAG, evaluation, harness engineering, and observability.
-
to Present
Voblakye
AI / Agent Engineer
Ongoing software and AI engineering work across internal initiatives.
-
to
TendencIA
AI & Automation Developer
Built WhatsApp and Telegram bots, customer-support chatbots, and internal automations using third-party LLM APIs. For AIRE, validated and integrated AI-backed flows and owned most of the Flutter and Supabase frontend.
Contact
Let’s build AI systems that hold up beyond the demo.
Open to remote AI / Agent Engineer roles across US and European time zones.
Based in Guadalajara, Mexico, with EST and CET overlap. If you are hiring for agent orchestration, grounded RAG, evaluation, reliability, or AI engineering infrastructure, I would like to hear about the role.
Also open to selected agentic-systems consulting and build collaborations.