Our Portfolio

Built pipelines that automatically extract data from incoming client documents—tax forms, statements, receipts—and file them into a structured, searchable storage system, cutting down manual data entry and document handling.

Trained on IRS publications and the firm's own past guidance, this assistant handles routine questions like "can I deduct X" from clients directly, freeing up billable hours for higher-value work.

Designed and built custom software tailored to individual client workflows and reporting needs, replacing generic off-the-shelf tools with solutions that fit how the firm actually operates.

Built an AI agent that reviews incoming client inquiries, gathers key case details, and routes each matter to the right attorney or practice group automatically, cutting down the manual work of screening leads.

Structured a firm's case files, filings, and correspondence into a knowledge graph linking matters, parties, and documents, giving attorneys a single place to search and trace case history and precedent.

Deployed a system that captures billable work as it happens, applies matter-specific billing rules, and generates client invoices automatically, reducing the manual effort of tracking and preparing bills.

Built an assistant that pulls quotes from multiple carriers, normalizes coverage terms and pricing, and presents clients with a clear side-by-side comparison, cutting down the manual work of assembling and explaining options.

Deployed an agent that gathers key details from inbound inquiries, scores them against the agency's qualification criteria, and routes ready-to-quote leads straight to agents, reducing time spent on unqualified prospects.

Built a system that requests, collects, and organizes the documentation underwriters need for each application, flagging missing items automatically and shortening the time from application to bound policy.

Deployed a computer-vision system that inspects parts and assemblies in real time, flagging defects the moment they occur so issues get caught before products move further down the line.

Built a monitoring system that analyzes equipment sensor data to predict failures before they occur, letting maintenance teams intervene on their own schedule instead of reacting to unplanned downtime.

Built a pipeline that reads incoming RFQs, matches specs against pricing and capacity data, and drafts quotes automatically, cutting the turnaround time between a customer's request and a firm quote.

Built an assistant that reviews project specs and drawings alongside historical job costs to draft estimates automatically, giving estimators a strong starting point instead of building each bid from scratch.

Deployed a system that turns field inspection notes and photos into structured, code-compliant reports automatically, reducing the time crews spend on paperwork after every site visit.

Built a system that optimizes crew scheduling, job dispatch, and material delivery routes against real-time constraints, cutting down idle time and last-minute rescheduling across jobs.

Deployed an agent that manages appointment booking, rescheduling, and reminders across phone and messaging channels, cutting down front-desk time spent on routine scheduling calls.

Structured a practice's clinical notes and patient records into a knowledge graph linking visits, diagnoses, and treatments, giving providers a fast way to trace a patient's history and surface relevant context.

Built an assistant that compiles visit and treatment data into properly coded, formatted insurance claims before submission, reducing rejections and the manual work of claim preparation.