The General Intelligence Company of New York moves with the quiet confidence of a team that knows the real story isn’t the headline but the machinery humming underneath it. Founded in Jan 2025 by Andrew Pignanelli and Abhishyant Khare after meeting inside South Park Commons, the company set up shop at 245 Lexington Ave with a question that felt closer to a challenge: if intelligence can compound faster than cost, why shouldn’t 1 person run a company built for 1,000. That question became Cofounder, an AI chief-of-staff that studies a business the way veteran operators study cash flow, linking into systems, forming memory, and acting with the steadiness of someone who already understands what matters and what never did.
Inside the NYC office, the story turns sharper. A support agent that replies, codes, evaluates, and moves changes toward production without waiting for human hands. A codebase that is 95%+ AI-generated with a target of 100%. A 3-layer memory system that tracks core habits, working context, and long-term knowledge through a living graph that grows as fast as the business feeding it. A dual-agent setup where a memory agent works nonstop while a real-time agent handles conversations, workflows, and 500+ step tasks that would exhaust even seasoned operators.
Cofounder’s public launch on Sep 8 pushed the company into the feed with speed, pulling in thousands of users in the first week. Benchmarks followed: 64.9 on Ruler-QA for retrieval accuracy and 15.8 on test-time learning for long-range recommendations, placing the system at the front of agentic memory tech. By Dec 7, the company closed an $8.7M seed led by Union Square Ventures with Acrew Capital, Compound, Scott Belsky, Yohei Nakajima, Agent Fund, Mute Ventures, basement.studio, John Phamous, Matthew Parkhurst, Sarah Chieng, Jacky Huang, and The House Fund joining in. Combined with the $2M pre-seed led by Michael Dempsey at Compound and Asad Khaliq at Acrew Capital, the team crossed $10M raised within 12 months of founding.
The capital backs a sharp wager. Andrew Pignanelli and Abhishyant Khare plan to show a software company run entirely by agents in 2026, using Flows to automate product, engineering, GTM, support, and ops as if departments were instruments in a single arrangement. If they pull it off, the idea of millions of new companies built by lean teams or solo founders shifts from theory to expectation, and the 400M-business global market starts to look like open territory for agent-native operators.
Startups Startup Funding Early Stage Venture Capital Seed Round AI Automation Agentic AI AI Agents Technology Innovation Tech Ecosystem Startup Ecosystem
