fin 254

I build backend and full-stack systems that pair distributed-systems engineering with AI-native features like embedding-based semantic search and LLM-powered chatbots. At NCBA Group (Carduka / AuctionsDuka), I designed a RAG pipeline with LangChain and OpenAI/Anthropic LLM APIs grounded in internal company data.

I also implemented semantic vector search using pgvector and vector databases like Pinecone/Weaviate, plus embedding generation and indexing pipelines to turn unstructured business data into retrieval-ready vectors.

On the backend side, I’ve exposed AI chatbot and vector search as REST APIs inside existing Java Spring Boot microservices, using gRPC, event-driven workflows, and optimized PostgreSQL schemas to reduce latency. I’ve built high-availability systems for real-time financial transactions, approvals, and vendor integrations, with CI/CD and monitoring via Azure Monitor and Application Insights.

Before that, I led enterprise iOS fintech development at Power Financial Wellness, bridging React Native and Swift, delivering offline-first workflows with SQLite and CoreData, and integrating secure APIs using WSO2 authentication. Earlier, I worked on Android fintech (MyCashi) and backend distributed financial transaction platforms with Kafka and OAuth2/JWT-secured REST APIs.