Harry W.

I partner with startup founders and product owners to build production-grade AI systems, fast. I'm a co-founder and full-stack architect who has built entire B2B and SaaS products from scratch, from the initial idea to scalable, production-ready systems. I specialise in turning complex business problems into structured and scalable databases, automated workflows and robust RAG pipelines. My systems are built to enterprise standards. As an AI Lead for a $70T+ investor network (FAIRR), I built a RAG pipeline that achieved 95% accuracy and replaced a manual analyst team. I bring that same rigour to your startup, ensuring we move fast without accumulating technical debt. --- Key Projects (Challenge → Solution → Result): 1. AI Automation for $70T+ Investor Network (FAIRR Initiative) - Challenge: Manual ESG analysis was slow, costly, and difficult to scale, requiring 3 full-time analysts. - Solution: Architected and built a full RAG pipeline with a multi-LLM workflow (OpenAI, Anthropic, Google) to automate the entire equity assessment process. - Result: Achieved 95% accuracy on industrial ESG analysis, reduced costs by over 70%, and proved the viability of enterprise-scale AI automation. 2. AI News Platform (Meridius Labs - Co-founder) - Challenge: Financial professionals needed a reliable, automated market briefing from hundreds of daily news stories. - Solution: As co-founder, I built the entire platform from scratch - proving I understand founder priorities: speed, capital efficiency, and user retention. I engineered a Human-in-the-Loop AI system with a 17-step modular workflow to analyse, summarise, and deliver briefings. - Result: The platform serves 950+ subscribers 24/7 on a lean £200/month operational cost - proving scalable, capital-efficient architecture. 3. Core Product Architecture (Oslo VC-Backed Startup) - Challenge: The startup needed a scalable, enterprise-grade architecture for its core BI product. - Solution: Architected the complete system: Google API ingestion → complex data storage → backend AI workflow → frontend agent delivery. - Result: Delivered a robust, production-ready architecture that helped the company secure its Series A funding. --- "World-class problem-solving skillset. Deep understanding of LLMs, AI and tools to solve complicated tasks. But the best part: A very friendly, interested and pro-active person." - Nicolai, CEO Paldo --- Why Startups Work With Me: Founder Perspective: I've built and scaled my own B2B SaaS product from scratch - I've navigated pricing pivots, infrastructure decisions, and the pressure of delivering features that move revenue metrics. Enterprise Rigor: You get systems-level thinking, tested at a $70T+ network, adapted for startup speed. I architect for scale from day one, not as a refactor six months later. Capital Efficiency: I build lean, scalable systems that maximise runway (e.g., £200/month infrastructure serving 950+ daily users). Strategic Partnership: I take ownership of your business outcomes. I join weekly standups, help shape your product roadmap, and think beyond the current sprint to what unlocks your next funding round. --- Technical Expertise: Intelligent Document Processing: RAG pipelines, multi-LLM orchestration (OpenAI, Anthropic, Google), vector databases, agentic workflows, prompt engineering Scalable Data Infrastructure: Neo4j knowledge graphs, PostgreSQL, Redis caching, Supabase, AWS (Glue, Lambda), GCP Full-Stack Development: Python (FastAPI), Celery (multi-queue task processing), React, HTMX, Docker, Kubernetes --- My Philosophy: A Researcher Who Builds As a Masters Natural Sciences graduate, my approach to problem-solving is rooted in first-principles thinking. I'm a self-taught systems architect who understands your business context before writing a line of code. I partner with founders who need more than just a developer - they need a strategic partner who will take ownership of the outcome. --- Let's Connect I offer a free 30-minute system architecture audit to help you assess your MVP approach. Message me to schedule a call. (Based in London/Asia, I keep UK hours for seamless collaboration.)