Fractional CTO
Senior engineering leadership embedded in your company — without the full-time hire, the equity negotiation, or the six-month ramp. Architecture, AI lab, ADR process, team mentorship. Mountain Time. Full US overlap.
What this covers
Not a consultant who shows up with a slide deck. Embedded work — async-first, with regular syncs. The things that fall through the cracks when your best engineer is also acting as tech lead, hiring manager, and architect at once.
Every non-trivial technical decision documented, contextualized, and stored. Future hires can read the reasoning, not just the outcome. Eliminates the "why did we do it this way?" tax.
Privacy-aware routing layers, fallback chains, per-workload model policies, semantic caching. The unglamorous infrastructure that makes production AI actually reliable.
Eval harnesses built from your real historical data before you ship. Not bolted on after the first incident. CI-ready, versioned, owned by the team.
LangGraph, tool use, human-in-the-loop escalation, state management. Architecture for agents that need to actually run in production — not just demos.
Technical interview design, take-home calibration, offer framing. Mentoring mid-level engineers on system design, distributed systems patterns, and production discipline.
Impartial evaluation of infra, model providers, observability stacks, and SaaS. No referral incentives. The recommendation your team gets is the one I'd make for my own money.
How it works
I read your codebase, your infra setup, your existing docs. I interview your key engineers and product leads. I produce a written technical assessment — what's working, what's load-bearing but fragile, what needs an ADR before you hire the next person. This audit is the start of the engagement, not a sales deliverable.
Weekly architecture reviews. Available in your Slack for async questions. PR reviews on the high-stakes decisions. Regular 1:1s with your tech lead and senior engineers. I show up in your standups when it matters. I don't show up when it doesn't.
Every decision gets documented. Every recommendation comes with rationale, alternatives considered, and the risks of the path not taken. When this engagement ends, your team owns the architecture — not me. The ADR library, the eval harness, the runbooks: all yours.
Most engagements run 6–12 months. Some extend. Some end when you hire a full-time CTO — in which case I help you interview them, brief them, and transition properly. Either outcome is a success. I'm not optimizing for contract length.
Who this is for
I work with a small number of companies at a time. Being honest about fit saves everyone time.
Engagement options
For teams that need a trusted technical voice without full embedding. Best for early-stage or as a trial before deeper engagement.
Full fractional engagement. Architecture governance, AI lab leadership, team mentorship, hiring bar. The work that actually moves the needle.
For companies building AI as a core product capability. I run your internal AI lab — prototyping, evals, routing, multi-agent design.
"Iván owns the system design for internal workflows — ADR process, service boundaries, async patterns, and data contracts. Separately runs the AI lab. It's the kind of embedded technical leadership that's hard to replicate with a full-time hire at our stage."
— Engineering lead, EqualsTrue · Ongoing engagement, 2yr+
Describe where you are and what's breaking. I'll tell you directly if it's a fit — and if not, I'll point you somewhere useful. No pitch deck. No discovery form.
Hermosillo, Sonora, MX · GMT−7 Mountain Time · Full US timezone overlap · Reply within 24 business hours.