Case studies
How the work was scoped, built, and shipped. Client engagements first, then Soltech's own product, then the R&D behind the methods.
Client engagements
- IFA Fund Report Pipeline: Automating Venture Capital Consolidation
Five-stage AI pipeline for a regulated UK IFA practice. Reads incoming venture fund manager PDFs, classifies every page, extracts portfolio data, and produces a structured brief the adviser can review without opening a single PDF.
- Signal Monitor: Commercial Intelligence for the Events Sector
Automated competitive intelligence for Momentum Works. Monitors 30+ UK events and exhibitions companies across website, RSS, LinkedIn, careers boards, and regulatory feeds; uses AI classification to pick out what actually matters, and delivers a branded brief and PDF to the client's own distribution lists every Friday, run from a self-serve web portal.
Product
- Setfolio: A Multi-Tenant .NET SaaS, Shipped With an AI-Assisted Process
A multi-tenant .NET 10 SaaS taken from first requirements to live on Azure in days, using a documented AI-assisted process that survived a mid-build change of tooling. Live and awaiting its first real users.
Lab
Personal projects, built off the clock. This is where the pipeline and classification methods used in the client work above were first worked out, on data I own.
- Changsta · an AI-assisted music platform, built with backend discipline
- TuneFinder · taste-driven release discovery, zero LLM calls per run
- MixLab · two-stage LLM mix curation from a Rekordbox library
- MixLab Playlist Completion · scored completions for a partial playlist
- Rekordbox Metadata Enrichment · confidence-scored lookups, delta XML back
Have a system that needs shipping?
Tell me what you are building and where it is stuck. I reply within one business day.