Projects
A working portfolio of experiments and projects from Three Eagles Forge Studio. This page is designed for engineers, founders, and hiring managers who want to see how ideas move from concept to shipped artifacts.
Each card below represents a real system in various stages of maturity—from R&D cycles to production-ready tools—with enough detail to understand the architecture, tradeoffs, and learning behind it.
VeilMark is an AI-authorship detection platform that combines locally hosted machine learning models with lightweight surfaces like a Discord bot, FastAPI service, and browser extension. It’s designed for community owners and builders who need a reliable signal on when content is likely AI-generated, without turning moderation into a black box.
Tech stack
VeilMark Discord Bot - Sample result for text analyzed by the VeilMark bot.
VeilMark Browser Extention - Checking the validity of an image.
The Domain Discovery Engine (DDE) is a multi-stage scoring pipeline that identifies high-quality brandable domains by combining availability checks, SEO signals, linguistic heuristics, and LLM-based evaluation. It provides a structured, repeatable workflow for identifying purchase candidates rather than relying on intuition or manual searching.
Tech stack
Representative weekly run summary showing top-20 candidates, availability, and scoring metrics.
csvMend and related utilities are targeted tools for cleaning and transforming CSV files, aimed at analysts, indie hackers, and builders who need quick, repeatable scripts rather than full-blown data platforms.
Tech stack
csvMend CSV feature overview showing key options.
Agentic Experiments is an ongoing R&D effort focused on testing multi-agent patterns, revision loops, and stateful workflows. The goal is to understand when agentic orchestration meaningfully improves output quality over single-prompt LLM calls, and how these patterns can support real 3EF products like PostStream, DDE scoring, and future decision-support tools.
Tech stack
Writing & deeper dives (coming soon)
Over time, this section will link to longer-form posts that unpack the architecture, tradeoffs, and results behind each project—case studies for VeilMark, DDE, csvClean, and the agentic experiments.
For now, the cards above are the best place to explore how I approach building indie-scale tools: concrete systems, constrained resources, and deliberate learning loops.
