Laid-Off Engineers’ OpenExecutive Puts Eight AI Executives Into One Open-Source Project

Laid-Off Engineers’ OpenExecutive Puts Eight AI Executives Into One Open-Source Project

N
News Editor
2026-08-27 10:08:07
OpenExecutive, an open-source project from startup Sente Labs, has drawn attention after a GitHub link reached the Hacker News homepage and scored 563 points on Aug. 27. The project is built around eight AI agents designed to take on executive functions, including strategy, finance, legal, operations, product, marketing, HR and board communications. According to the project page, the system can pull context from a vector database, store decisions in SQLite and be called through Slack, Telegram, Discord, email or the command line. The project gained extra attention after the uploader, GrumpySciGuy, said in the Hacker News comment thread that the team behind it had been laid off. He wrote that after losing their jobs, the engineers started a company and decided that while tools such as Copilot and Claude Code were changing software development, they wanted to change management. At the same time, the repository’s README does not mention that backstory, and the repo itself was created on June 11, suggesting development had already been underway for about two and a half months. OpenExecutive also includes 10 built-in executive “persona” files, nine of them based on public talks and earnings calls by living CEOs including Jensen Huang, Mark Zuckerberg, Satya Nadella and Dario Amodei. As of Aug. 27, the repo had 795 GitHub stars, 45 forks and 12 open issues.

OpenExecutive, an open-source project built around AI executives, has climbed onto the Hacker News front page with 563 points. The submission was posted under the headline: “CEO laid off engineers and replaced them with AI, engineers built an open-source AI CEO.”

The project centers on eight AI agents designed to handle executive roles, from strategy and finance to board communications. As of Aug. 27, the GitHub repository had 795 stars and 45 forks.

Uploader said the team had been laid off

The repository was uploaded by an account named GrumpySciGuy. In the comment thread under the Hacker News post, the user said the laid-off group was their own engineering team. He wrote that after the team was let go, they decided to start a company: “We think Copilot and Claude Code are changing development. We want to change management.”

That backstory does not appear in the project’s GitHub README. The report also noted that the repo was created on June 11, indicating the project had already been in development for roughly two and a half months. When asked which company had cut the team, the uploader did not answer.

Eight AI executives with multiple interfaces

Sente Labs lists OpenExecutive as a flagship product on its website. According to the company’s description, the system can organize signals accumulated overnight into decisions, coordinate specialists across departments, monitor outside events and produce board presentations.

The eight agents map to a chief strategy officer, chief financial officer, chief human resources officer, chief legal officer, chief operating officer, chief marketing officer, chief product officer and director of board communications. Before responding, each one retrieves relevant context from a vector database. Decision records are stored in SQLite, and the system can be accessed through Slack, Telegram, Discord, email or the command line.

Ten built-in “tone” files, nine based on living CEOs

One of the more unusual parts of the repo is the personas/builtin directory. It contains 10 built-in tone files, nine of which are drawn from real, living executives. The list includes NVIDIA’s Jensen Huang, Meta’s Mark Zuckerberg, Microsoft’s Satya Nadella, Google’s Sundar Pichai, Apple’s Tim Cook, Amazon’s Andy Jassy, Airbnb’s Brian Chesky, Stripe’s Patrick Collison and Anthropic’s Dario Amodei.

These files are written as behavioral instructions. The Dario Amodei profile tells the AI to use probabilistic language naturally and to be “willing to seriously discuss scenarios that most business leaders consider too speculative to address.” The Jensen Huang profile tells it to explain complex ideas through concrete imagery tied to manufacturing processes, physics and the geometry of markets. The files also specify words to avoid, including “ecosystem effects” and “platform leverage.”

The public demo runs on fictional companies

The repo’s test material also shows where the project’s limits are. It includes three sample companies, all fictional: Halcyon Motors, Meridian Petroleum and Tandem Robotics.

Each company comes with a memory.json file containing preset decision records. In the oil company example, a decision dated April 8, 2026 says the company would hedge its third-quarter crude position with Brent. The reason given is: “A refinery without hedging is just betting on a war we can’t predict.” In the robotics example, the company shifts to a Robots-as-a-Service model, charging by robot work hours instead of selling the full machine outright.

The report said the public demo runs on that scripted setup and is not actually managing a real company.

Community testing is still underway

The project currently has 12 open issues, and the community is still testing how far it can really go. The comment section quickly turned playful. One user wrote, “This CEO just laid off the entire C-suite, thanks,” while another said layoffs could finally start from management.

OpenExecutive is an open-source project developed by Sente Labs. The biggest unanswered question remains the same: which company laid off the team behind it. So far, that detail has not been disclosed.

This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan.
30

Disclaimer:

The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.

Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.