ABOUT NANANAMI

A small team. Senior credentials.

Nananami is a Seattle-headquartered AI operations consultancy. We are paid on the value we create for our clients, measured in their own systems. Architecture, sales, and client-facing from Seattle. Board of Advisors with European PhDs and Swiss industry credentials.

Why we exist

Most operations leak value in places nobody is looking: overtime committed in a plan weeks before payroll notices, coverage approved by chat against limits nobody checks, revenue that is earned but takes months to be recognized and collected. The data to see it usually already exists. What is missing is the system that surfaces it in time to act.

Nananami exists to build that system, and to be paid only when it works. We diagnose where value leaks in the client's own data, build the AI-powered tools and agents that capture it, measure the result in the client's own systems, and share in what we created. The client keeps the large majority. We think that is the only honest way to sell AI to an operation: on outcomes, not on hours or licenses.

Our founder spent years at Roche and Silicon Valley health-tech startups, shipping software and prediction models in environments where a claim has to be backed by evidence. That discipline carries over, together with the structured problem solving of the McKinsey tradition that our AgentMinds pipeline is built on: every number we put in front of a client is auditable in their own records, and every rule our systems enforce is one the client wrote.

We pair that with a Board of Advisors of three members who anchor distinct technical practice areas: distributed systems and program analysis, signal processing and sensor AI, edge AI and embedded systems. International research and patent records that hold up to diligence. And for medtech and digital health teams that need to ship AI features under FDA rules, we keep a specialized advisory lane where that same background is directly useful.

The team

AV

Abel Villca

Founder & Architect

Engineer with background at Roche and Silicon Valley health-tech startups. Founder of Nananami. Designed AgentForge, AgentMedReg, and AgentMinds. Architect on every engagement and personally accountable for the value measured on each one. Based in Seattle, WA.

LinkedIn →

Board of Advisors

Three advisors anchor distinct technical practice areas: a computer scientist trained in Geneva and Innsbruck, a laser physicist from the Minsk school, and an electronics engineer working in Swiss industry. Each reviews architecture and signs off on engagements matched to their expertise.

Alex Villazón

Systems Architecture

MSc and PhD in Computer Science, University of Geneva; postdoctoral research at the University of Innsbruck (Austria) and Università della Svizzera italiana. Two decades in distributed systems, dynamic program analysis, cloud computing and IoT across European research and industry. Senior reviewer on systems architecture for production AI engagements.

Omar Ormachea

Signal Processing & Sensor AI

PhD in Laser Physics from Belarusian State University in Minsk, the Soviet school of laser physics, where he was a senior researcher in non-linear optics. 70+ publications in laser physics, photonics and spectroscopy; twice national science and technology laureate; member of the Bolivian National Academy of Sciences. Senior reviewer on signal-processing pipelines and sensor-data architectures.

IT

Ivan Terceros

Edge AI & Embedded Systems

R&D electronics engineer at Daxtro SA in Switzerland. Specialist in IoT, embedded systems, PCB design, low-power energy harvesting and LPWAN (LoRa, Sigfox). National first prize in science and technology for portable laser spectroscopy. Bridges Nananami's engagements to production hardware and edge AI.

Proprietary multi-agent architectures

Three patterns we developed, ship, and operate. Each is in production on real engagements.

8 AGENTS

AgentForge

Sequential software development pipeline from feature request to deployed, monitored production code. Orchestrator routes through Strategist, Analyst, Architect, Developer, QA, DevOps, Monitor.

AgentMedReg

FDA-regulated software development pipeline. Bookend pattern with Regulatory Strategist opening and closing. Covers IEC 62304, ISO 14971, 21 CFR 820.

AgentMinds

Strategic problem solving applying structured-analysis frameworks. Engagement Lead, Structurer, Analyst, Challenger with quality gates between stages.

How we work

Paid on value, measured in your systems

No hourly billing. Our fee is a share of the value we create, and the measurement lives in your own records, not our estimates.

Documentation is part of the deliverable

Runbook, ADRs, recorded handoff. Not an afterthought.

Multi-agent in production is the bar

Tutorials are not the same as shipping. We run what we sell.

Plain English, no jargon walls

Domain terms when they are correct. No invented words to sound smart.

Other ways we engage

Value-based engagements are our main line. Two specialized lanes exist for work where a scoped, defined engagement fits better.

MEDTECH & DIGITAL HEALTH

FDA-Validation Advisory

Architecture review, implementation and retainer for teams shipping AI features under IEC 62304 / ISO 14971 / 21 CFR 820. Scoped engagements with named deliverables.

Explore →

ARCHITECTURE ADVISORY

Senior outside read of your AI system

For AI, DSP and ML architecture questions that are not a value engagement and not FDA work. Reviewed by the founder and the relevant advisor. Quoted on request after a short call.

Book a call →

Let's find what's leaking.

A short diagnosis in your own numbers. If there is value to capture, you see it before you commit to anything.

Book a diagnosis