About

Member of Technical Staff · London, UK · Hybrid

Driving enterprise integrations and partnerships for Doubleword's high-throughput AI inference API, helping to drove adoption for open source LLMs for background tasks for use cases spanning coding, evaluation, synthetic data generation and persistent memory for agentic systems.

Founder

Find what's next, now. A freshness-first semantic search engine for developers and AI agents, focused on real-time indexing and hybrid retrieval. I architect the full stack independently, from crawling and data infrastructure through to ranking and retrieval logic.

Stack Rust, Go, Python, AWS, Kubernetes, Terraform

• Refactored the core data processing pipeline, cutting runtime across 1.5 billion records from two months to three days.
• Designed modular crawling and ingestion infrastructure on AWS EC2/EKS, deployed via Terraform and Helm.
• Implemented query expansion and hybrid vector/keyword retrieval with controlled ranking logic and performance instrumentation.
• Built detailed analytics over complex evaluation and click-analysis datasets.
• Defined AI guardrails and deterministic filtering to remove and downweight low-quality or inappropriate content, using stop words, prompt templates, query routing, and domain/content-based ranking.
• Expanded index coverage across news, blogs, codebases, and academic papers.

Software Engineer · Remote

A 0-to-1 role building the core matching and data infrastructure that connects startups with investors. I designed reliable systems across ingestion, retrieval, ranking, and reporting. I joined as a venture capital analyst (Aug 2025) advising global startups on fundraising before moving into engineering.

Stack Go, Python, TypeScript, Postgres/DuckDB, vector DBs, Docker, Terraform, Grafana, Prometheus

• Built a production investor-startup matching engine generating roughly 90,000 matches for 80+ fundraising clients.
• Implemented hybrid retrieval pipelines combining vector search, embedding similarity, and deterministic scoring.
• Created an investor-to-investor matcher using a custom GPU-accelerated Node2Vec implementation in PyTorch Geometric.
• Built ETL workflows to scrape, clean, enrich, and validate large datasets across Parquet, Postgres, and BigQuery, with QA in LangSmith.
• Built agentic onboarding flows with tool use to write structured answers to Postgres and trigger OCR document extraction.
• Designed evaluation loops to measure ranking precision, reduce duplication, and improve match quality.
• Built robust retries, resumability, and idempotency for long-running pipelines.

Vice President, Partnerships · San Francisco Bay Area · Hybrid

A go-to-market and technical role at ChipHub, an AI-enabled SaaS platform in the NVIDIA Inception program serving the semiconductor industry. I built internal tooling and a web datasheet search feature while owning OEM partnerships.

• Built a datasheet search feature spanning millions of datasheets, enhancing an Azure-deployed document RAG pipeline in Go and Python and cutting procurement research time by roughly 70%.
• Developed internal prospecting tools for ICP discovery, replacing £3,600/year in third-party tooling.
• Pitched to and managed relationships with global OEM clients.
• Sourced and onboarded UK-based frontend developers, creating PRDs, roadmap prioritisation, and customer feedback loops.

Venture Capital Analyst · Dubai, UAE · Hybrid

Sourced and evaluated startups across MENA for Spartech Ventures Fund II, covering F&B tech, healthtech, and embedded finance. Built a process connecting the region's leading family offices with deal flow, in partnership with the Hub71 accelerator.

• Built and deployed a web-scraping and matching engine using embedding-based similarity scoring (BERT, all-MiniLM-v6) and automated ranking.
• Extended the matching engine for a Hub71 partnership, connecting 100+ family offices to startup deal flow.
• Sourced targets, wrote investment memos and one-pagers, and took companies through due diligence.
• Produced research shaping the fund's technology thesis, used directly in DDQs with a sovereign fund.

Product Lead · London, UK

Product and technical lead at an Oxford University Innovation-backed future-of-work startup building blockchain credential infrastructure on Sidetree (Microsoft x Ethereum). I progressed from Research Associate to Product Associate to Product Lead.

• Managed a junior developer building an ML platform converting unstructured text into the standardised ESCO skills taxonomy with spaCy NER and BERT.
• Built prototypes in Streamlit and React, and led ethnographic feedback sessions with students.
• Pitched the platform to university registrars across the UK.
• Represented Certie at London Tech Week 2023 and the Gibraltar Ministry of Finance.

IT Lead / Tutoring Supervisor · Nottingham, UK

• Managed tech administration across four school programmes in Nottingham, overseeing examinations and data collection used in impact assessments critical to funding bids.
• Tutored GCSE English pupils from disadvantaged backgrounds, supporting their academic progress and confidence.