Skip to content

About EganForge

AI engineering, practiced in-house.

Egan Forge is an independent development studio for businesses and entrepreneurs who need useful AI inside real workflows. The same operating system used to deliver client work also supports our own SaaS products, so recommendations are tested against systems we have to maintain—not presentation slides.

Stylized hammer and anvil overlooking West Virginia mountain ridges
2026
Founded
West Virginia
Headquarters
12
SaaS products available
3
Clear service paths

The short version

One operator, systems designed to last

Egan Forge is a solo-operator studio. Josh Egan leads discovery, architecture, implementation, and handoff while using purpose-built agents for research, testing, documentation, and routine operations.

The automation is not the product by itself. The product is a workflow that works when inputs are messy, APIs fail, and a person needs to take control. That is why the studio emphasizes bounded scope, observable systems, and plain-English ownership at handoff.

The SaaS catalog is the proving ground. Each available product has to be shipped, supported, measured, and either maintained or taken out of promotion. That same discipline carries into client development work.

How it's structured

T0

Human oversight

CEO reviews, approves high-impact decisions

T1

Master agents

Planning, health monitoring, cost control, and delivery checks

T2

Delivery systems

Client workflows, SaaS products, integrations, and support

T3

Task agents

Individual workers inside each system

High-impact changes require explicit human approval. Independent model review and an audit trail are used where the cost of a wrong answer justifies the added friction.

What we believe

Principles, not platitudes

01

Ship small, useful tools

Every product solves one problem. No Swiss-army-knife apps with 40 features nobody uses. One tool, one job, done well.

02

Free tiers should actually work

If the free tier is too limited to be useful, it's just a trial with extra steps. Ours are sized so most users never need to upgrade.

03

AI should save time, not create work

Nobody wants to spend 20 minutes crafting the perfect prompt. Our tools take simple inputs and handle the complexity behind the scenes.

04

Be honest about what AI can't do

AI makes mistakes. We say so on every product page, in our terms, and in our AI transparency statement. Verify before you rely.

What we build with

Practical systems, selected for the job

Interfaces

  • Next.js and TypeScript web applications
  • Clear operator views and approval paths
  • Responsive, accessible customer experiences

AI systems

  • Models selected by task and constraint
  • Structured outputs, evaluations, and fallbacks
  • Human review where consequences require it

Operations

  • Python services, workers, and integrations
  • Observable deployments and failure paths
  • Documentation, ownership, and handoff