Sethfounder / builder
← Projects
Open Source AI System

Research

A completely free AI agent designed to give people access to advanced AI workflows - without expensive subscriptions, fragmented tooling, or walled gardens.

Why This Exists
I believe powerful AI systems should be accessible to more people - not locked behind increasingly fragmented software stacks and compounding subscription fees.
Why are AI tools fragmented?
Every major AI lab ships its own interface. ChatGPT, Claude, Gemini, Perplexity - each one solves part of the problem, none solve all of it.
Why do users switch between apps constantly?
Because no single tool handles research, writing, planning, execution, and learning in one place. Users lose context every time they switch.
Why are advanced workflows expensive?
Premium tiers across multiple tools add up fast. Most people pay for three subscriptions and still feel underserved.
Why is AI still hard to use effectively?
The interface, not the intelligence, is the bottleneck. Most people never get past basic prompting because the tools aren't designed for deep workflows.
The Current Landscape

Right now, a typical power user jumps between:

ChatGPTClaudeGeminiPerplexityCopilotNotion AITrelloObsidianLinearDozens more...
→ Context lost with every switch. Subscriptions stacking. Intelligence fragmented.
The Vision
A free AI system capable of helping users think, learn, plan, research, create, organize, and execute - from a single place, without fragmentation, without paywalls.
Free
Core workflows accessible to everyone, not just paying subscribers
Open
Open source architecture. Community-driven. Transparent development.
Unified
One system for thinking, learning, planning, writing, and executing.
Core Features (Planned)
Research Mode
Research topics deeply
Gather and rank sources
Organize findings
Generate summaries
Build knowledge bases
Learning Mode
Create study plans
Track learning progress
Generate explanations
Build skill roadmaps
Adaptive curriculum
Productivity Mode
Task management
Planning systems
Goal tracking
Calendar integration
Execution frameworks
Writing Mode
Draft articles and posts
Generate outlines
Organize ideas
Content systems
Long-form research
Project Mode
Manage projects end-to-end
Track milestones
Generate documentation
Create roadmaps
Progress reporting
Agent Mode
Long-running tasks
Workflow automation
Multi-step reasoning
Background execution
Future integrations
Public Build Log
→ Building in public. Updated as work progresses.
June 2026
Exploring architecture options
Studying agent frameworks
Mapping possible workflows
Benchmarking existing tools
Aug 2026
Building first prototype
Testing memory systems
Designing the interface layer
Early workflow experiments
Oct 2026
Initial user testing
Refining agent loop
Adding research mode
Performance benchmarking
Future →
Public beta
Open source release
Community contribution system
Plugin architecture
Why It Matters

The current AI landscape rewards fragmentation. Every company wants you in their ecosystem. Every new tool adds another subscription. The person who benefits least from this is the one actually trying to get work done.

The Open AI Agent is not a product play. It's a belief that the intelligence layer should be open - the same way Linux made operating systems accessible, and the same way open source made software development possible for everyone.

Advanced AI workflows shouldn't require multiple paid subscriptions
Open source AI systems can compete with closed commercial products
The interface is the bottleneck - not the intelligence
A unified AI layer compounds in value over time
Interested in contributing or following the build?
→ Open source when ready · build log updated publicly
Get in Touch →