Fairwind is an operations and customer-management platform for trade contractors, connecting customer and site history, sales intake, estimates, dispatch, work orders, crew execution, and service and asset-care workflows for home-service teams.
For general contractors, Fairwind supports project administration, trade-partner coordination, subcontract changes, progress claims, closeout, and deadline notifications. Shared purchasing workflows connect supplier catalogs, requests for quotation, purchase orders, receiving, and inventory-related activity.
Financial workflows cover project costs, supplier bills, payables, receivables, and billing, with owner-focused reconciliation and workspace data export. Role-based access and shared records help teams follow customer work, field activity, commitments, materials, and costs in one workspace. Capabilities are at different stages of development and release readiness.
Explore FairwindMoon Frontier explores how vector search can keep pace with data that changes over time. Its native search engine is designed to support inserts, updates, and removals without requiring a full index rebuild for every routine change.
Benchmarking and visualization tools help engineers inspect search behavior and compare approximate retrieval with an exhaustive-search baseline on representative data. Rather than promise universal performance gains, Moon Frontier helps teams evaluate the quality and cost trade-offs for their own workloads.
Under active development; not currently offered as a hosted service or production-ready database.
Anansi is an open-source Model Context Protocol (MCP) server that gives AI assistants a single, typed GraphQL interface to application data. Instead of exposing a separate tool for every API endpoint, agents can inspect the available schema and compose precise queries, including requests for related data, in a single call.
Built-in validation and configurable query limits help keep requests focused and predictable. Anansi currently includes an in-memory sample dataset and is designed to be extended to connect with real databases, REST APIs, and other services.
It's a practical exploration of how to make existing systems easier for AI agents to work with.
Lyrics to Learn turns song lyrics into language-learning material. Learners can search for songs through the Genius API or paste lyrics directly, then explore the lyrics section by section.
AI can generate vocabulary explanations and contextual practice exercises, including answers and explanations, to help learners study language through music.
Main Software designed and built Tohi, a full-stack web platform that brings personal finance, nutrition, sleep, and activity tracking together in one experience. Users can manage their goals through a unified dashboard.
A modular API connects the application to services such as Plaid for financial data, Stripe for subscriptions, and AI features for nutrition support.
Built with Next.js, TypeScript, and PostgreSQL, Tohi provides a flexible foundation for expanding features as user needs grow.
