TinyMacro
Enterprise-ready multi-agent orchestrator system for automating knowledge work with AI. Features vectorized data traversal and runbook-guided worker agents. Built with TypeScript, React, AI-SDK, and PGvector. Note: this is a boilerplate, and has been used for specific implementations for past clientele.
MyHRBuddy
AI-powered Applicant Tracking System with natural language search, resume parsing, and Zapier integration. Built with Next.js, Prisma, PostgreSQL, and MCP server integration to fully automate HR workflows.
FileBuddy
CLI tool for local file conversions. Converts images, downloads YouTube audio, and processes files for my own personal use cases. I built this tool to streamline the process of adding mp3 files to my underwater headphones for triathlon training. Built with Node.js, FFmpeg, and Sharp for high-performance media processing.
Drive Organizer
AI-powered Google Drive image organizer using Gemini Vision API. Automatically categorizes and organizes images based on content analysis. Features React frontend, FastAPI backend, and Redis for metadata storage. [Project still in Progress]
Enterprise-Scaled Context Engine for AI Implementations
Designed a hybrid AI context solution to solve data navigation challenges in enterprise AI deployments. Conducted comprehensive competitor analysis across RAG and Knowledge Graph solutions, performed technical feasibility research, and developed detailed product specifications including cost modeling and risk mitigation strategies. Proposed architecture combines vectorized data indexing with semantic chunking to improve precision@k metrics for company-specific terminology.
Technical specification document covering competitive analysis, feasibility research, and cost modeling: View Business Case PDF. Please reach out to me at frederickcaseyhousand@gmail.com for the password.
AI Image Organization and Compression Algorithm
Led product development for a company-wide image database organization system. Defined requirements and scope for AI-powered categorization using AWS Rekognition. Launched to immediate adoption with ~100k+ API uses in first month, exceeding launch targets. Designed and oversaw development of web portal with Smart Search functionality for marketing and sales teams to efficiently navigate 100GB+ image library.
Document built for colleagues, demonstrating smart search and AI categorization system: View System Guide PDF. Reach out to me at frederickcaseyhousand@gmail.com for the password.
AI-Powered MarTech in 2025
Identified opportunity to improve marketing engagement through AI-powered personalization and automation. Built verticalized AI tools that automated CRM workflows within HubSpot, deployed customizable automation with Facebook client funnels, and developed an AI agent using customer data to send personalized messages through social media. The solution improved clickthrough rates by approximately 20% and doubled response rates, driving measurable adoption across the marketing team.
