Building software people rely on.
Every project has pulled me deeper into engineering. I started by building interfaces, but curiosity led me to backend systems, APIs, databases, and the architecture behind products that people rely on every day.
Where I've Made an Impact.
- 01
Unscroll
Founder · Full-stack EngineerBuilt Unscroll from the ground up—an AI customer support platform that helps businesses answer customer questions using their own knowledge base. Designed the backend architecture, real-time conversations, asynchronous document processing, semantic search, authentication, and the customer dashboard.TypeScriptNext.jsNestJSPostgreSQLpgvectorRedisBullMQ
2026Founder - 02
Nepsix Technology
Full-stack Engineer · MentorBuilt and maintained production features across multiple company products, from backend APIs to responsive user interfaces. Mentored aspiring developers through hands-on training, helping them build real-world projects and strengthen their engineering fundamentals.TypeScriptNext.jsNestJSExpressPostgreSQLMongoDBPrisma
2024 - PresentProfessional - 03
TheSuzerainDesk™
Frontend EngineerWorked closely with the team to bring their brand and product to life by building a fast, responsive, and polished marketing website that reflects their vision and delivers a great user experience.Next.jsReactTypeScriptTailwind CSS
2025Client Work - 04
Eggrow Africa
Frontend EngineerPartnered with the team to transform design concepts into a production-ready website, creating a responsive, accessible, and modern experience that communicates the company's mission.Next.jsReactTypeScriptTailwind CSS
2026Client Work
Beyond the work featured here, I've built personal projects, client work, and products that live in private repositories or are covered by NDAs.
Explore all projects→
codewithemmaI build products
people actually use.
I started programming in 2021, teaching myself Python with no idea where it would eventually lead. A year later I discovered frontend development, got hooked on React, and spent countless hours learning how to build interfaces that felt fast, intuitive, and reliable.
In 2024, I began working professionally as a software engineer. Since then, I've collaborated with talented engineers, designers, and founders to build products used by real people—from AI-powered platforms to education technology and startup websites.
Today I work across the stack, turning ideas into production software with technologies like TypeScript, Next.js, NestJS, PostgreSQL, and React. Whether I'm designing an API, building a dashboard, or shipping a new feature, I enjoy solving problems that make products better for the people using them..
Beyond building software, mentoring has become one of the most rewarding parts of my career. I've helped 20+ aspiring developers learn frontend development, and seeing many of them grow into confident engineers is something I'm just as proud of as the products I've shipped.
I'm still learning every day, still building, and always excited to work with teams creating products that make a real impact.
- Based in
- Port Harcourt, NG
- Focus
- Product engineering
- Stack
- TS · Node · MongoDB
WHAT I do.
- 01
I Build
End-to-end products that solve real problems.
- 02
I Architect
Systems that stay reliable as they scale.
- 03
I Optimize
Performance, developer experience, and maintainable code.
- 04
I Mentor
Helping developers become confident builders.
- TypeScript
- JavaScript
- Python
- SQL
- React
- Next.js
- Tailwind CSS
- Mantine UI
- NestJS
- Node.js
- Express.js
- Prisma
- PostgreSQL
- MongoDB
- Redis
- pgvector
- Docker
- BullMQ
- Git
- Vercel
UNSCROLLAI support that feels native.
the problem.
Small teams often can't afford dedicated support staff, yet customers still expect immediate answers. Traditional chat widgets either rely on scripted FAQs or expensive live agents. The challenge was building an AI assistant that could answer accurately from company knowledge while allowing seamless escalation to a human when needed.
the approach.
I built an embeddable AI support platform powered by Retrieval-Augmented Generation (RAG). Company documents are processed asynchronously, chunked, embedded, and indexed with pgvector for semantic search. Incoming questions retrieve relevant context before streaming responses from the language model. Every conversation can be handed over to a human agent without losing context.A three-stage pipeline — candidate generation from a vector store, a lightweight cross-encoder for re-ranking, and an on-device attention model that decides when to taper. Everything streams; nothing blocks the first paint.
the result.
A complete customer support platform with a configurable widget, organization dashboard, knowledge base management, conversation history, AI/human support modes, streaming responses, and API-key-based embedding for any website. The architecture was designed to swap AI providers with minimal changes while remaining production-ready for real-world deployments.
Let's build something
worth shipping.
I enjoy working with founders, startups, and growing teams to turn ideas into reliable, production-ready software.