Nil SakariyaEngineering at the edge of AI.
Full-stack engineer & founding partner at DNJ TechX, CTO at Yrambio Solutions. I build AI products for healthcare, legaltech, and education — from zero to production.
About
A builder who treats software like an instrument — tuned, opinionated, and made to perform.
I've spent the last six years shipping production systems across healthcare, legaltech, and education — from RAG pipelines that summarise medical records to LLM-powered decision tools used by clinicians daily.
My day-to-day: making AI products feel inevitable rather than impressive. Strong opinions about latency budgets, weak opinions about frameworks. I believe the best AI products are the ones users forget are AI — they just work.
Selected work
Three products.
One ambition: useful AI.
Each of these started as a real domain problem with a stakeholder on the line, not as a demo. Built for production, not for pitches.
- 01 / 03AI Product · Healthcare
GastroAGI
An AI-powered clinical reasoning engine for gastroenterologists. Processes patient histories, lab results, and endoscopy reports to surface differential diagnoses, suggest next-step investigations, and auto-generate structured clinical notes — reducing documentation time by 60% in pilot deployments.
Next.jsLLMRAGHealthcareFHIR - 02 / 03AI Product · LegalTech
LexaMedAI
Bridges the gap between legal compliance and medical practice. Ingests medico-legal case files, extracts key entities, cross-references regulatory databases, and produces audit-ready summaries — enabling law firms and hospital compliance teams to cut review cycles from weeks to hours.
ReactNLPAgentic AILegalTech - 03 / 03AI Product · Education
MBBSAGI
An adaptive study companion for MBBS students that generates exam-style questions, explains concepts with layered depth, and tracks weak areas across 19 subjects. Uses spaced repetition algorithms combined with LLM-generated explanations to boost retention and first-attempt pass rates.
Vector DBRAGAdaptive LearningEdTech
Founder
Building companies that ship AI products people depend on.
Not just writing code — shaping product direction, hiring engineers, closing enterprise deals, and making sure the thing works at 3 AM when a doctor needs it.
DNJ TechX
Founding PartnerAn AI products studio building production tools that clinicians, lawyers, and students actually use daily. We don't build demos — we build systems that handle real patient data, real case files, and real exam pressure.
Yrambio Solutions LLP
Chief Technology OfficerAn AI engineering consultancy that converts client ideas into shipped software. We take companies from 'we want to use AI' to 'our AI system processes 10,000 requests/day' — with clean architecture, proper observability, and zero vendor lock-in.
Tech Stack
Tools I reach for when it matters.
Not a bingo card — these are the tools I've shipped production systems with.
Languages
Frontend
Backend
AI / ML
Infrastructure
Philosophy
How I think about building things.
Structure over speed
I don't ship fast and fix later. I ship structured and iterate confidently. Every system I build has clear boundaries, typed contracts, and a migration path.
AI where it adds leverage
Not every problem needs a model. I use AI surgically — where it genuinely reduces human effort or unlocks capabilities that weren't possible before.
Own the outcome
I don't hand off half-built prototypes. From architecture to deployment to the first user complaint — I stay in the loop until the product works in the real world.
Latency is a feature
Users don't care about your model size. They care about how fast the answer appears. I obsess over perceived performance, streaming responses, and edge caching.
Trajectory
From freelance to founder.
Founding Partner
DNJ TechX
Co-founded an AI products studio shipping production tools across healthcare, legaltech, and education. Architected the core platform powering GastroAGI, LexaMedAI, and MBBSAGI. Lead a distributed engineering team, own technical strategy, and drive product from zero-to-one.
Chief Technology Officer
Yrambio Solutions LLP
Run the engineering org for an AI consultancy that turns client ideas into shipped software. Designed the internal automation framework that reduced project delivery timelines by 40%. Oversee architecture reviews, hiring, and client technical relationships.
Software Engineer (Full Stack)
Freelance / Independent
Built and shipped 15+ production web applications for startups and SMBs. Specialized in React/Next.js frontends with Node.js backends. Developed a reputation for clean architecture, fast delivery, and zero hand-holding post-launch.