01
Clarity while the path is still open
Turn a promising but messy idea into one defined problem, one near-term outcome, and a sequence the team can defend.
Yenson Umaña · AI architecture partner for startups
AI architecture for startups.
From AI ambition to production clarity.
Startups are where the new AI world reaches users first. I help founders and CTOs turn that speed into clear decisions, production-ready architecture, and a path their team can execute.
Senior AI Solution Architect supporting Microsoft for Startups globally via Accenture.
The startup reality
Innovation moves fast. Architecture decisions compound faster.
Early AI companies rarely lack ideas. They lack time, certainty, and spare engineering capacity. Models change, product assumptions move, and every shortcut can become tomorrow's platform.
My biggest lever is pattern recognition from working with startup teams at this exact stage: I help the team see what matters now, what can wait, and what must be true before production.
02
A production-ready architecture
Make the model, data, retrieval, evaluation, security, cost, and human-review decisions before they become expensive rework.
03
A senior partner beside the team
Work directly with someone who can move between founder priorities and engineering tradeoffs, with no sales layer or junior handoff.
What startup advisory actually entails
I join when the team has momentum, but no map.
Early companies rarely arrive with clean requirements. They arrive with customer signals, investor pressure, a changing product, a working demo, and a dozen decisions nobody has time to untangle.
I work alongside the founder and engineers until the next production decision is clear, documented, and owned by the team.
- 01
- Many promising ideas
- One prioritized AI bet
- 02
- A prototype that works once
- An evaluated production path
- 03
- Model and platform noise
- Explicit architecture decisions
- 04
- Context held by the founder
- A roadmap the team can execute
Built for startup speed
A clear path from idea to production, without a six-month consulting layer.
Focused engagements that meet the company where it is: choosing the bet, designing the system, or helping a growing team operate AI well. Scope follows a free discovery call.
01
1-2 weeks
AI Direction Sprint
Turn a crowded field of AI possibilities into one clear first bet and a path the team can execute.
Best for: Founders and CTOs with strong market insight, several plausible AI directions, and no shared way to choose what deserves to ship first.
- Use-case portfolio scored by value, feasibility, and risk
- Startup readiness and capability gap assessment
- 90-day roadmap with owners and decision gates
- Founder and technical-team readout
02
2-4 weeks
AI Architecture Sprint
Turn one consequential AI idea or prototype into a production-ready system plan.
Best for: Startup teams moving an agent, RAG system, or AI-native product capability toward production without a senior AI architect in-house.
- Target architecture and architecture decision records
- Human-AI boundaries, guardrails, and escalation paths
- Evaluation, observability, latency, and cost plan
- Phased delivery roadmap and technical handoff
03
3-6 weeks
AI Team Enablement
Give a growing startup shared AI practices before founder knowledge and one-off experiments become delivery bottlenecks.
Best for: Scale-ups with pilots or production AI, but inconsistent evaluation, unclear ownership, or teams learning the same lessons separately.
- Founder and engineering-lead alignment
- Right-sized governance and production playbook
- Team workflows, evaluation patterns, and review standards
- Office hours, pilot support, and adoption measures
After the sprint
Fractional AI Architect
Ongoing, after an initial sprint
Senior technical judgment for a startup that needs an experienced AI architecture partner before it is ready to hire that role full time.
- Founder and architecture working sessions
- Model, vendor, and platform decisions
- Technical risk and delivery oversight
- Team unblockers and production reviews
Startup work
I know what the room feels like before the path is obvious.
Startup architecture means making consequential decisions with an incomplete map. The work below shows how I turn urgency and constraints into a production path the team can own.
Anonymized startup engagement
A full production migration completed in one week, with no visible downtime.
- full migration
- 7 days
- visible downtime
- 0 min
- team handoff
- Full
Startup migration · Production path
7 days- 01
Urgent founder objective
- 02
Target architecture
- 03
Agent-assisted execution
- 04
Acceptance gates
Ambiguous migration
Team-owned system
Constraint
A US-based Y Combinator startup needed to move its production stack from Vercel and Google Cloud to Azure without disrupting customers.
Intervention
I turned an urgent founder objective into the target architecture, acceptance gates, and an agent-assisted execution path the engineering team could follow.
Outcome
The full stack moved in seven days with zero customer-visible downtime, plus a runbook and operating approach the team could continue using.
Named client · Amplification Of Potential
A live-event AI system designed around the conversations it must never create.
- p95 latency target
- < 2.5s
- inference budget / event
- < $1
- PII in model payload
- Zero
AOP Beacon · Decision flow
GuardrailedPhase objective
01Guardrail policy
02Few-shot + fallback
03Private client render
04Attendee selections
Safe conversation prompt
Constraint
AOP needed personalized prompts in under 2.5 seconds without sending attendee names to the model or allowing sensitive topics into an early event phase.
Intervention
I designed the phased experience, prompt architecture, client-approved few-shot bank, explicit topic guardrails, privacy boundary, and deterministic fallback path.
Designed result
A reusable architecture for six active tables and up to 40 attendees, with a sub-$1 event inference budget and zero PII in the model payload by construction.
Founder / operator proof
I also live with the decisions after launch.
Founder and operator
Presencia Loyalty
Built and operate a live wallet-based loyalty SaaS, translating product strategy into a system businesses and customers use.
Connected product experience
Junior Rodríguez × Presencia
Designed an NFC-enabled painting experience that joins a physical object, its story, and a shareable digital journey.
How I work
Founder context, architecture, and execution stay in the same room.
Every engagement produces decisions the technical team can execute without losing the founder's product context. Tools and models are selected after the operating problem is clear.
- 01
Frame the decision
Clarify the business outcome, baseline, users, constraints, and what must be true for the initiative to deserve investment.
- 02
Make tradeoffs explicit
Compare viable approaches across quality, latency, cost, security, maintainability, and organizational readiness.
- 03
De-risk with evidence
Use focused prototypes, evaluations, and failure-mode reviews where they resolve an important unknown - not as theater.
- 04
Transfer the capability
Leave decisions, standards, and operating knowledge with your team so the work compounds after the engagement ends.
Model and vendor independence
Human review where consequences demand it
Evaluation before automation confidence
Ownership transferred to your team
Your startup architecture partner
I help early teams find the path before the playbook exists.
My strongest body of work is with founders and lean engineering teams making product, model, cloud, evaluation, and cost decisions while the product itself is still moving.
Through Microsoft for Startups, I see the patterns between a promising demo and a production system across different teams. That pattern recognition is the leverage I bring to your table. You work directly with me, with no sales layer or junior handoff.
2022
Production AI before ChatGPT
Joined OneReach.ai in October 2022 and began building conversational AI and orchestration systems at production scale.
2023-24
One-person platform ownership
Owned roadmap, tooling, and architecture for Wind River's Engineering Excellence function before moving into full-stack platform engineering.
Today
Startup architecture at the decision table
Work with founders and lean engineering teams on AI architecture, agent systems, evaluation, productionization, and technical direction through Microsoft for Startups.
Current independent work is limited to non-conflicting engagements and does not imply endorsement by Microsoft, Accenture, or any current or former employer.
Start a conversationBefore we talk
Useful answers, without the sales call.
Do you implement the systems you design?+
I use prototypes and technical validation when they reduce a material risk, and I can support an internal team through delivery. The core offer is senior strategy, architecture, and adoption - not open-ended outsourced development.
What kind of company is the best fit?+
Early and growth-stage startups with a founder or CTO close to the decision, a product or engineering team ready to execute, and an AI initiative important enough to shape the company.
How does this work alongside your current role?+
I accept a limited number of non-conflicting engagements, each subject to a conflict review. Client work is independent and does not imply endorsement by Microsoft, Accenture, or any current or former employer.
Is training available on its own?+
Yes, when it is tied to a defined operating change. I design role-specific executive and team sessions around your systems, governance, and adoption goals rather than generic AI literacy presentations.
Can you work with a distributed or international team?+
Yes. I work from Costa Rica with teams globally in English, using a mix of live working sessions and documented asynchronous decisions.
How do you handle confidentiality?+
A mutual NDA is available on request. Engagement boundaries, data access, model usage, IP ownership, and retention expectations are agreed before sensitive material is shared.
Free 30-minute discovery
Bring the AI problem your startup cannot yet turn into a path.
We will clarify the real constraint, the next production decision, and whether working together would move the team forward fast enough to justify the investment.