We build platforms that turn complex data into something useful. Knowledge graphs, interactive maps, real-time systems, and AI-driven tools across legal, financial market, enforcement, and historical research. If it involves large datasets and hard problems, we are probably already working on it.
Novel procedural scene description language and runtime rendering engine. Real-time world morphing driven by live public blockchain state, tracking six-figure on-chain asset mutations per block via a custom ledger parser. Procedural decimation system achieving 80%+ bandwidth reduction across 5,000 concurrent users.
Acquired by international technology group14 interconnected platforms built on a 10 million+ entity knowledge graph with 2.66 million resolved relationships. Automated ingestion pipelines processing 127,000 legislative items, 435,000 court cases, and 84,000 statistical records. Blockchain-anchored verification engine running continuous checks against primary government sources.
justiceindex.ukReference record for tokenised assets: 7,757 instruments across nine wrapper structures, each with the issuer, the wrapper, redemption terms, dividend treatment and where the holder ranks if the issuer fails. Sourced and dated. A scheduled monitoring run checks every source and commits what changed, and a Failure Files series documents where products and venues have broken.
xrade.comData pipeline ingesting and reconciling 13,000+ Domesday settlements from the Palmer/Hull collection (University of Hull), the National Archives' Manorial Documents Register, and Victoria County History entries across 12 counties. Geospatial rendering engine, search algorithms, and statistical analysis tools.
manors.org.ukCustom static site generator producing 30 localised sites from a single data-driven pipeline. Templating engine, automated cross-site linking algorithms, programmatic SEO generation, and bulk deployment orchestration. 3,700+ pages built and deployed from structured data.
solent.newsThe company's trading name and research studio. Applied AI on proprietary NVIDIA infrastructure: frontier model evaluation, agentic workflow orchestration, and reliability testing across real-world deployments, alongside data and monitoring tools for tokenised markets. Systematic documentation of failure modes and architectural insights. Member of UK AI.
botresearch.aiFull-stack Next.js application with Supabase backend. Automated source enrichment pipelines ingesting from eight external APIs (Historic England, TNA Discovery, Wikidata, BHO). Interactive geospatial rendering, network visualisation, relational data model linking 115+ manors across lordship chains, and real-time search.
sheldonmanors.org.ukIngesting, structuring, and serving large datasets from multiple sources. Real-time state tracking across hundreds of thousands of records per cycle.
Entity resolution, relationship mapping, and graph-based data models. 10 million+ entity graphs with millions of cross-referenced connections.
Frontier model deployment on proprietary NVIDIA infrastructure, agentic workflow design, reliability testing, and applied AI research.
Procedural generation, runtime rendering, and live data processing for thousands of concurrent users. Performance optimisation on consumer hardware.
Smart contract integration, blockchain-anchored data integrity, real-time chain monitoring, and anti-fraud validation systems.
Multi-layer spatial analysis, interactive mapping, risk assessment overlays, and location-based data aggregation from government and institutional sources.
Enforcement data and research: digital asset tracing, cross-jurisdictional analysis, and the structure of enforcement law and process. Informed by practitioner experience and postgraduate research.
Automated pipelines connecting government, institutional, and commercial data sources. Report generation, continuous verification, and multi-source reconciliation.
End-to-end platform development from prototype to production. Next.js, React, Node.js, PostgreSQL, cloud infrastructure, and deployment automation.
Sheldon Research Ltd, trading as Bot Research, is a UK software research and development company founded in 2022. We build data-driven platforms across multiple research domains, from legal infrastructure and enforcement intelligence to AI systems and geospatial analysis.
Our work combines large-scale data processing, knowledge graph architecture, and interactive mapping. We have a particular affinity for maps. From real-time 3D environments and procedural rendering to historical cartography and multi-layer geospatial risk analysis, most of our platforms have a spatial dimension because location is often the best way to make complex data intelligible.
The common thread across everything we build is software that processes, structures, and presents information in ways that did not previously exist. Same team, same tools, same research-led approach, applied wherever the problem is interesting.
Director: Morgan Sheldon. Company number 14206452. Registered in England and Wales. Trading as Bot Research.