Looking for a job

henry ernest

@henryernest

Message

AI Full Stack Engineer at BPOSeats who built a Mistral-powered RAG system that improved retrieval accuracy by 40%.

Uganda

Message

At BPOSeats, I built a Mistral 7B-powered RAG system with HuggingFace embeddings and Weaviate, improving retrieval accuracy by 40%. I also developed a multi-agent system for creating and publishing technical content, reducing the manual content cycle time by 70%.

My work spans AI products and full-stack platforms, from MCP developer tooling and LLM-based recruitment at BPOSeats to fraud detection models at CIM Innovations. I also created Owlvora Learning, a multilingual education platform designed for low-bandwidth environments, and have fine-tuned models for Swahili text classification and multilingual speech recognition.

Experience

Work history, roles, and key accomplishments

Current

Senior Full Stack Engineer

Current

BPOSeats

Jul 2023 - Present (3 years 4 months)

Built scalable RAG systems and multi-agent architectures, improving retrieval accuracy by 40% and reducing manual content cycle time by 70%. Led cross-functional teams and engineered AI-driven platforms for recruitment, matchmaking, and HR compliance.

Django Vue.Js NodeJS React Next.js TypeScript Express Sequelize CubeJS OpenAI LLaMA 3.1 Pinecone RAG MCP Servers AI Microservices Crew AI Lang Chain Hugging Face Weaviate LlamaIndex PyTorch

Software Engineering Fellow

Headstarter

Jul 2024 - Sep 2024 (2 months)

Rapidly prototyped and deployed 5 AI projects within 5 weeks, reducing development cycles by 30%. Built AI customer support agents and SaaS products, scaling web apps to over 200+ active users in the first month.

React Next.js TypeScript Firebase Vercel OpenAI Pinecone Stripe LLaMA 3.1 Groq API Python Flask Machine Learning RAG PyTorch scikit learn

Full Stack Engineer

CIM Innovations

Mar 2020 - Jul 2023 (3 years 4 months)

Developed high-performance video recording tools and global payment gateway solutions, integrating multiple regional payment providers. Engineered machine learning models for fraud detection, reducing fraudulent activities by 35%.

NodeJS Express TypeScript Django Flask React PostgreSQL Docker Python Machine Learning PyTorch scikit learn CubeJS Flutterwave PayPal Alipay Adyen

Software Engineer

Hackers Yacht

Jun 2018 - Jun 2019 (1 year)

Developed a scalable online university noticeboard platform using Python, React, and React Native, improving event participation by 40%. Led backend optimization efforts, reducing server costs by 20% and accelerating data retrieval speed by 30%.

TypeScript NodeJS React React Native Python Flask Django JavaScript