FauxRun
A multi-tenant AI platform that analyzes client and project context to generate structured proposals, outreach, and writing — backed by a 12,000+ skill autocomplete library and a Chrome extension that inserts AI content directly into web forms.
Challenge
Teams writing proposals, outreach, and client-facing content repeatedly retype the same context — a client's needs, an organization's writing style, the skills relevant to a pitch — with no shared memory across writers or across submissions in online forms.
Approach
FauxRun combines a multi-provider LLM engine (Gemini, OpenRouter, and others) with structured, type-safe output to analyze content for needs, challenges, and themes, then generate proposals, summaries, and outreach in an organization's own style, using few-shot examples and a 12,000+ entry skills autocomplete library for consistency. A Chrome extension fetches an organization's styles and profiles from the FastAPI backend and inserts AI-generated content directly into online forms, cutting the copy-paste step out entirely. True multi-tenant isolation with role-based access (system admin → org admin → user) keeps each organization's data and defaults separate, backed by MongoDB Atlas and Celery/Redis for async processing.
Outcome
Organizations get consistent, on-brand proposals and outreach generated from reusable context instead of rewritten from scratch each time, with token usage and generation history tracked per user and per org, and a browser extension that removes the last manual step of getting AI content into the forms that actually matter.