Monchis
Monchis
Architectural redesign of a React Native + microservices mobile platform on AWS. Stability and performance lift that restored user retention and engineering momentum — the methodology that earned the muv reassignment.
Head of Engineering
Building AI-first engineering teams
AI-first engineering & architectural transformations · Claude Code, AWS & GCP · Ex-Mercado Libre · Ex-IADB.
Contact me on LinkedIn 6+ years scaling teams
Currently
Two products. One playbook.
Two products inside itti's ecosystem. Same playbook, different scale: stabilize, redesign, ship, then turn the same approach into team-level AI adoption.
- 01 Aligned engineering, product and design across a matrix org — negotiated priorities with business stakeholders, engaged senior leadership on technical strategy and quarterly planning.
- 02 Designed and shipped micro-apps architecture that decoupled mobile deploys, unblocking independent team velocity.
- 03 Introduced AI-assisted development (Claude Code, Cursor, Superpowers) across a 12-person team — faster cycles, less repetitive overhead.
- 04 Championed code review culture, observability tooling and incident response that brought reliability to production-grade levels.
- 05 Cloud infrastructure on AWS + GCP — compliance, scalability and uptime as engineering org-wide standards.
Mercado Libre · 2022 – 2024
Financial infrastructure trusted by 100+ internal teams
Project Leader at MercadoPago. Led teams of 8 and 12 engineers across design, delivery, and execution — collaborating directly with technical leads on architecture decisions while owning team performance and roadmap.
- Financial Core
Transfers & Debts engine
Co-architected high-stakes Golang processing tools running daily multi-million-dollar batches at MercadoPago's operational core.
- Internal Platform
One Source
MercadoPago's main visibility console — single pane pulling from every API across the ecosystem. Daily contact with most engineering teams in the company.
- Developer Experience
Frontend Framework
Drove adoption and evolution of a widely-used internal frontend framework — thousands of projects depended on it, other frameworks built on top of it.
Inter-American Development Bank
Delivered engineering for multiple projects at one of the world's most influential multilateral development banks — managing cross-functional teams across engineering, design, product, and policy.
- 60+
- People managed across engineering, design, product, policy.
- 30%
- Time-to-market reduction via microfrontend architecture (Angular + Vue).
- C-Level
- Direct technical-strategy reporting. Roadmap influence in a large matrix org.
Engaged initially as a full-time consultant, then continued as a flexible contractor leading internal projects — a structure that allowed concurrent engagement with Mercado Libre from May 2022 onward.
AI-First Engineering
Where AI changes how teams ship
Practical, deliverable, repeatable. From org-wide Claude Code rollouts to side projects pushing agentic workflows past the obvious — and the management practice required to make it real.
Speaking on AI-first engineering — beyond the demo
Talks aimed at engineering and management teams. Not just "how to become AI-first" — the harder management-scope problems nobody is talking about: incident playbooks, code review semantics, secret management, harness composition.
Harness & Agentic Workflows
How to compose tools, memory, and the agent loop that actually ships product. Patterns from real rollouts at itti.
Management in an AI-first world
Code review when AI is in the room. Incident recovery when the bug was AI-authored. Secret management, harness integration, agentic workflow design — the management problems the demos never cover.
Claude Code Rollout · itti
Driving AI-first transformation across thousands of engineers
Member of itti's core AI team (<40 people) gesting the AI-first transition for an engineering org of thousands. My contribution went beyond the obvious.
$ claude-code init --org itti
→ bootstrapping engineering org
→ aligning code review for AI-generated PRs
→ incident playbooks · secrets · harness
✓ ready for production
▍- Code review semantics when AI authored the diff. Trust signals, blast radius, rollback playbooks.
- Incident recovery when the bug is AI-authored — debugging conventions, escalation paths, post-mortem framing.
- Secret management + harness integration so agents can act without leaking credentials or context.
- Spec-Driven Development (SDD) + deep agents — composable workflows the rest of the org can adopt without rewriting.
Side Project · D&D AI Companion
An agentic RPG platform pushing the architecture further
Multi-tenant agentic platform for tabletop RPG. Helps DMs and players create characters, build full campaigns, translate adventures across rulesets — without reading thousands of pages of source material.
- Architecture
Graph-per-tenant, graph-per-campaign
Each campaign is its own graph. Lore, characters, and rule references stay isolated and queryable.
- Interoperability
Translation layer between rulesets
Bridges D&D 5e, Pathfinder, Vampire: the Masquerade and more — non-compatible schemas reconciled in one runtime.
- Orchestration
Multi-provider LLM routing
Claude + GPT in the same pipeline. Pick the right model per task — reasoning, narration, world-building, character builder.
- Harness
Custom tools + harness in-code and as product
Custom tools, in-code harness for trusted ops, and a product-level harness for end-user agents to compose without leaking context.
- Agent-as-a-Service
Beyond SaaS — agents you can subscribe to
Not just a SaaS app: every campaign gets its own scheduled, persistent agent. Identity, memory, and orchestration are first-class.
- Scheduled workflows
Agentic loops on a clock
Long-running, scheduled workflows that prep next session, reconcile state, and surface plot threads the DM forgot.
- Introduction to Model Context Protocol Completed
- Introduction to Agent Skills Completed
- Building with the Claude API In progress
- Claude 101 Completed
- Claude Code 101 Completed
- Claude Code in Action Completed
- Model Context Protocol: Advanced Topics In progress
- Introduction to Subagents Completed
Beyond Engineering
My Hobbies
Where curiosity goes after work hours.
Hobbies · D&D
Dungeons & Dragons
Dungeon Master with friends. Organization, improvisation, and empowering people's creativity — the same instincts that translate back to engineering leadership. Also the seed of the D&D AI Companion side project.
- Role Dungeon Master · story-first
- Skills Organization, improvisation, empowering creativity
- Crossover Leadership rhythm — and the seed of a side project
Hobbies · Spirituality
Spirituality
Studious by nature, focused on non-duality. Applying it day-to-day to deconstruct, soften ego, and become a calmer, more complete person — better partner, better leader, better human.
“Less ego, more presence. Better partner, better leader, better human.”
Want to know more?
Best path is LinkedIn. I read everything, reply to anything technical, and answer recruiters who do their homework.