Portrait of Bernard Adala Wanyande Software (AI) Engineer at ING

I build AI that turns documents into data.

I'm Bernard Adala Wanyande, a software engineer in ING Wholesale Banking, working on AI-powered extraction use cases. I like the part where models meet production: APIs, pipelines, and the cloud infrastructure around them.

I sit somewhere between software engineering and data science. After an MSc in Computer Science (Data Science) at Leiden University and a stint building Generative AI tools at Unilever, I now build production AI systems at ING.

Day to day that means Java and Python services, LLM-based agents, and the DevOps work that gets them safely into production on Google Cloud.

  1. Sep 2026 – Present
    Current

    Software (AI) Engineer · ING

    Wholesale Banking

    Building AI-powered document extraction use cases. I develop APIs and services with Spring Boot and FastAPI, build agents on Google's Agent Platform, and ship to Google Cloud Run through automated DevOps pipelines.

    • Spring Boot
    • FastAPI
    • Google Agent Platform
    • Cloud Run
    • CI/CD
  2. Nov 2025 – Sep 2026

    Teaching Assistant, Quantum Computing · Q-Team

    THUAS, TU Delft & Leiden University

    Taught quantum computing fundamentals through lectures and lab sessions, contributed to curriculum development, and organised workshops across the university network.

    • Quantum Computing
    • Teaching
    • Workshops
  3. Feb 2025 – Sep 2025

    Data Scientist, Generative AI · Unilever

    Automated shelf-life prediction, cutting manual analysis time by 60%. Built an AI learning assistant that turns SharePoint content into learning paths, and a RAG chatbot over proprietary company data.

    • Python
    • LangChain
    • OpenAI
    • FastAPI
    • Azure DevOps
DutchMaster AI interface

DutchMaster AI

2026

Multi-model app for Dutch homework. Runs OpenAI, Anthropic and Gemini in parallel to analyse homework photos, explain grammar and decode idioms.

  • Next.js
  • FastAPI
  • OpenAI
  • Anthropic
  • Gemini
Enterprise RAG knowledge system

Enterprise RAG Knowledge System

2025

MSc thesis with Unilever: retrieval-augmented generation over internal documents using hybrid search and LLMs.

  • Python
  • LangChain
  • FAISS
  • Streamlit
YouTube dual-subtitle Chrome extension

YouTube Dual Subtitles

2025

Chrome extension that shows real-time translations above YouTube subtitles, built for language learners with the Gemini API.

  • JavaScript
  • Chrome Extension
  • Gemini
Recommender system visualisation

Hybrid E-Commerce Recommender

2025

Pair-based implicit recommender framed as classification over engineered user–item features, evaluated on Recall@10.

  • Python
  • LightGBM
  • Random Forest
  • SVD