About Intent Computing
Intent Computing, Inc. is a Chicago company. Grigore Rosu founded it in 2026 to put his ideas about proof-checked AI code into practice. Rosu's view is that tests only approximate what a person means, so computing needs proof as its check. He describes a model where a person approves a specification in their own words. AI then writes the code and a machine-checked proof that the code matches it.
The company's address is 180 N Stetson Avenue, 35th Floor, Suite 3547-B, in Chicago. Rosu is its founder, president and chief executive officer, and he is affiliated with the University of Illinois Urbana-Champaign. Zhixuan Huang is an executive officer and director, and Musab Alturki is the secretary. Rosu was scheduled to present the idea at the Fields Institute in Toronto on October 26, 2026.
Funding Rounds (SEC Form D)
| First sale | Filed with SEC | Amount sold | Investors | Avg. check | Terms | Filing |
|---|---|---|---|---|---|---|
| Oct 3, 2026 | Oct 23, 2026 | $62,500of $500K target | 4 | $15,625 | Simple Agreements for Future Equity (SAFEs)Rule 506(b) | Form D ↗ |
First sale is the date the first investor in the offering committed to invest. Companies must file Form D within 15 days of that date, so filing dates run later.
Why Investors Backed Intent Computing
Intent Computing, Inc. is a Chicago company founded in 2026 by Grigore Rosu. Rosu argues that AI works best where success can be checked, and that for software the check should be proof. His idea is that AI writes code together with a specification and a machine-checked proof. He founded the company to bring that idea into everyday use.
The problem it targets
Rosu's starting point is that a test only approximates what a person meant by a program. In his group's experiments, at least 15 percent of AI-written code that passed all tests could be proven wrong. His case is that proof is the missing check for computing, and that AI has now removed the last obstacle to using it.
What the product does
The approach Rosu lays out has AI generate code together with a formal specification and a machine-checked proof that links the two. The work runs over executable formal semantics of real programming languages. He reports that when AI also writes the specification and proof, it writes better code and solves more tasks. In his picture, programming languages fade from view once every answer carries a proof.
Progress so far
Rosu says his research group has settled two questions about matching logic, the logical foundation of the K Framework, with AI as a collaborator. One is the global completeness of the basic logic. The other is that no complete proof system can exist once fixpoints are added. His group had left both questions open for years.
We wrote this summary from public statements by Intent Computing and its investors. It is not investment advice.
Leadership Team
Executive officers and directors listed on Intent Computing's Form D filed Oct 23, 2026.