Full-Stack AI Engineer · Hyderabad, India
I build AI products, and the AI-powered tools companies run on
I design and ship web and mobile apps end to end with Next.js, React Native and Node, and lately I spend most of my time building AI systems: RAG pipelines, LLM integrations, and internal tools built with AI agents, in production use at real companies.
Currently: building AI internal tools for an ad agencyPreviously: Neshama (AI reflections app), client web & mobile builds
Open to conversations
- Name
- Praveen George Ryan
- in/praveen-george-ryan-a4598327b
- Based
- Hyderabad, India · IST
- Languages
- English · German · Hindi · Telugu
- TypeScript
- Python
- Next.js
- React
- React Native
- Expo
- Node.js
- Postgres
- pgvector
- Supabase
- MongoDB
- RAG Pipelines
- AI Agents
- Claude API
- OpenAI API
- Docker
- GCP
- AWS
- Vercel
Selected work
Shipped products, real users
6 projects shown.
More projects
Zionel Agency Website
Agency site with services, portfolio and case-study sections.
React Next.js TypeScript Swiper CSS
(opens in a new tab)Nafa Digital Marketing Website
Marketing site for a digital advertising agency specialising in Google Ads management.
React Next.js Tailwind CSS
(opens in a new tab)Animated Portfolio Website
Personal portfolio for a digital marketing professional, with a WebGL 3D scene and Framer Motion animation.
React Next.js Three.js Framer Motion Tailwind CSS
(opens in a new tab)Three.js / GSAP Website
Animated portfolio for a clinical research professional, built around a Three.js scene and GSAP motion.
React Vite Three.js GSAP Tailwind CSS
(opens in a new tab)Fresh to Cook
Online ordering site for fresh fish and seafood, with a menu, basket and sign-in.
React Vite
(opens in a new tab)Exceed Prep Academics and Sports
Course and enrolment site for a coaching academy combining academics with athletic training.
React Next.js TypeScript Tailwind CSS
(opens in a new tab)SmileMakers Dental Care
Dental practice site with appointment scheduling.
React Next.js Tailwind CSS MongoDB
(opens in a new tab)Women's Care Clinic
Clinic site with online appointment booking and patient registration.
React Next.js TypeScript Tailwind CSS
(opens in a new tab)Skin Vitality Medical Clinic
Lead-capture landing page for a medical aesthetics clinic, with a consultation request form.
React Next.js Tailwind CSS
(opens in a new tab)VU Global Consultancy
Site for a consultancy serving businesses and individuals.
React Next.js TypeScript Swiper
(opens in a new tab)
About
I build the whole thing: web and mobile front ends, the backend and data layer underneath, and the AI systems that make the product worth using.
I've been building software since October 2021. I started my career as an R&D scientist (2019–2021), and when a deep interest in technology pulled me into software, I brought the scientist's method with me: treat every idea as a hypothesis, change one variable at a time, measure the result, write everything down. In software it goes by hypothesis-driven development; for me it means small, measurable steps and products that ship rather than prototypes that stall.
Today that means RAG pipelines, LLM integrations and internal AI tools that real teams depend on daily: meeting intelligence, client memory, semantic search over a company's own knowledge. I build them with a heavily AI-assisted workflow: agents write a lot of the code, I direct the architecture, review every change and own what ships. I care most about the parts that decide whether people keep using a thing: latency, failure modes, and what happens after a wrong answer.
I'm focused on agentic AI systems and AI engineering.
Praveen George Ryan
Full-Stack AI Engineer
Hyderabad, India
Stack
- Web & Mobile
- Next.js · React · React Native · Expo · TypeScript · Tailwind
- Backend & Data
- Node.js · Express · Postgres · MongoDB · Supabase · REST / GraphQL
- AI & Infra
- LLM integration (Claude / OpenAI) · RAG + vector DBs · Docker · GCP · AWS · Vercel
Services
What I take on
Four kinds of work, each backed by something I have shipped. The case studies under each one show how.
- 01
AI integration & LLM systems
Retrieval, search and assistants built into products people already use, with the failure modes designed for up front rather than found in production.
- RAG pipelines and semantic search over a company's own data
- Assistants that act inside an interface, not just chat beside it
- Batch generation and model fallbacks to keep cost and latency in check
- Measuring retrieval quality before trusting it
- 02
Full-stack web apps
Web products built end to end: the data model, the API, the interface and the deploy.
- Typed APIs over Postgres or MongoDB
- Multi-tenant SaaS, user roles and two-factor sign-in
- Payments, pricing rules, transactional email and owner dashboards
- Integrations with CRMs, project tools and Google APIs
- 03
Mobile apps
Cross-platform React Native and Expo apps, from the first screen to store-ready builds.
- One codebase for iOS and Android
- In-app subscriptions with RevenueCat
- AI features served from the backend, so no provider keys ship inside the app
- Push notifications and background sync
Seen in
- 04
Migrations & rescue work
Taking over a codebase that is stuck, slow or locked into a platform: finding what is actually wrong, then fixing it in safe steps.
- Moving apps off a vendor platform onto your own cloud
- Database migrations, such as MySQL to PostgreSQL
- Security reviews that close data-access gaps
- Accessibility and Core Web Vitals work
Seen in
Not sure which of these fits? Tell me what you are building
Say hello
Let's build something
Hiring for an AI engineering role, or have a product that needs building? Tell me what you're working on.
- Based
- Hyderabad, India · IST
- Elsewhere
- GitHubLinkedInRésumé, PDF
Or send a message