Software, AI, and data systems

I build data systems, dashboards, and software for complex analysis. My work includes AI-assisted tools, language-model integrations, and agents that connect models to real data and workflows.

Litigation data support, apps, and dashboards

Data platforms, dashboards, and applications that help teams interrogate large sampling, exposure, and document datasets.

Service for legal teams

Large, document-heavy cases need reliable data infrastructure, whether or not an expert is testifying. I build that foundation as a separate service.

AI can accelerate the extraction of information from scanned lab reports and spreadsheets, as well as repetitive data-cleaning tasks. The results still need review by someone who understands the chemistry, sampling, and case context. That expertise is essential to reliable results.

Tell me about the matter

Data intake and standardization
Lab reports, spreadsheets, and monitoring data combined into one clean, documented dataset.
Dashboards
Maps, charts, and tables that let the legal team and experts explore the data themselves.
Gap analysis
What the data cannot yet support, and what would close the gap.
Recommended next steps
Sampling and analysis plans, additional analyses, and next steps.

Tools I have built for legal teams

Software, AI tools, integration, and agents

Applications and tools that use language models and machine learning, plus agents that connect models to real data and workflows.

  • Rataplat

    Live

    A recommender for boulderers. Climbs are tagged by style, climbers log what they send, and the model suggests climbs at an area that match their strengths and tastes, which matters at huge areas like Fontainebleau.

    Python, Recommender systems, NLP, Android

    Details on Rataplat

  • smartlyt

    Live

    An AI tool that groups PubMed search results into clusters of related papers, so a deep literature review starts from a few relevant islands instead of hundreds of unsorted abstracts.

    R, Shiny, TF-IDF, Hierarchical clustering

    Details on smartlyt

  • PubMed literature review and alerts

    Live

    A literature review workflow on top of Claude's PubMed connector. It pulls every study on a topic, screens out the noise, builds an evidence table, compares what the papers say, and sends recurring alerts when new work appears.

    Claude, MCP, PubMed

    Details on PubMed literature review and alerts

  • Timecard generator

    In development

    An agent that fills in a timecard from the work itself. It reads email, calendar, and Teams chat, matches the activity to projects, and drafts the hours, so you review a draft instead of reconstructing the week from memory.

    Python, Microsoft 365, LLM agents

    Details on Timecard generator

  • HR assistant

    In development

    A chat assistant that answers employee questions about HR policies, benefits, and procedures using the company's own documents.

    Python, Retrieval, LLM agents

    Details on HR assistant

  • NFL pick'em tools

    In development

    An MCP server that gives a language model live access to rating systems, market lines, and my pick history for pick'em strategy, plus a mobile app for win-total predictions and pick logging.

    Python, MCP, React Native, Expo

    Details on NFL pick'em tools