The Open Engine Company
The Open Engine Company builds open-source tools for AI software development. Its mission is to make AI free and to make software agents more reliable by enforcing independent validation, review, and repair workflows.
At a Glance
- Individual software developers and technical founders
- Startups building software with AI coding agents
- Engineering teams using Claude Code, Codex, or GitHub Copilot
- Organizations requiring repository-specific QA, review, and delivery controls
- +1 more
AI Tools by The Open Engine Company
(1)
zeroshot
QA Loop CLI for Coding Agents
Discussions
No discussions yet
Be the first to start a discussion about The Open Engine Company
Latest News
Opcore continued active development; the public repository recorded its 0.3.0 full rewrite and 292 commits.
VIS reported that The Open Engine Company secured a place in the national DNB 100 Pitches semifinal after Demo Day Vestlandet.
The Open Engine Company won the Demo Day pitch competition in Bergen and advanced to the national 100 Pitches final; the company also discussed its undisclosed recent capital raise.
Eivind Meyer appeared at a CTO Roundtable in Oslo to discuss robust AI development at scale and Zeroshot.
Products & Services
Free, MIT-licensed open-source CLI and cloud/self-hostable workflow engine for AI-assisted software development. It executes an explicit graph in which a worker implements a task, independent acceptance and code reviewers assess it in parallel, rejected work goes to a bounded repair loop, and accepted work can optionally be delivered through Git/CI.
Open-source code-intelligence and robustness layer for coding agents. Its installed hook checks current Git changes and reports the file, location, rule, and evidence requiring repair; it is designed to run alongside existing tests and linters rather than replace them.
Market Position
The company positions Zeroshot against the limitations of single-agent coding workflows and identifies Anthropic and OpenAI as major competitors because they can build similar capabilities internally. Its differentiation is an open-source, model-agnostic, mechanically enforced graph: the agent that writes code cannot approve its own work, and rejected work is routed through independent review and bounded repair. Opcore adds a lower-level, deterministic edit-validation layer that complements Zeroshot.
Leadership
Founders
Eivind Meyer
CEO and co-founder. He has a background in cybernetics from NTNU and is pursuing PhD research in self-driving cars at the Technical University of Munich; he and Michael Eichelbeck met while doing machine-learning PhD research there.
Michael Eichelbeck
CTO and co-founder. He met Eivind Meyer during machine-learning PhD research at the Technical University of Munich and is associated with research in machine learning, robotics, and safe reinforcement learning.
Olle Kronvall
CMO and co-founder. He has seven years of startup-growth experience, including seven years in startup/growth work at DNB, and leads The Open Engine's brand and growth.
Executive Team
Eivind Meyer
Chief Executive Officer (CEO)
Co-founder with NTNU cybernetics training and PhD research in self-driving cars/machine learning at the Technical University of Munich.
Michael Eichelbeck
Chief Technology Officer (CTO)
Co-founder and machine-learning researcher associated with the Technical University of Munich; leads the technical/product-engineering direction.
Founding Story
The founders started the company after using AI agents on a roughly one-million-line software project and experiencing what they described as total chaos: agents could not autonomously build and validate a large project reliably. They set out to build an orchestration and verification layer in which agents review and validate one another from different perspectives, with the longer-term vision of open, scalable software manufacturing.
Business Model
Revenue Model
The core Zeroshot CLI and Opcore are free and open source. The company also offers Zeroshot Cloud as a managed workflow service; the cloud documentation describes connecting GitHub and using customers' own model API keys. The local/self-hosted products can use existing Claude Code, Codex, or Copilot subscriptions/logins.
Pricing Tiers
Zeroshot is MIT licensed and can run locally or on infrastructure controlled by the user; Opcore is open source and Apache-2.0 licensed.
Pricing shown on the Zeroshot website; models run on the customer's own API keys and the charge is per extra parallel run, not per seat.
Target Markets
- Individual software developers and technical founders
- Startups building software with AI coding agents
- Engineering teams using Claude Code, Codex, or GitHub Copilot
- Organizations requiring repository-specific QA, review, and delivery controls
- Researchers running reproducible, agent-assisted experiments
- Implementing software features through reviewed pull requests
- Fixing bugs with acceptance and code-review gates
- Risky changes and dependency upgrades requiring security, blast-radius, or adversarial review
- End-to-end and UI testing of agent-written software
- Repository-wide refactors and migrations
- Continuous validation of coding-agent edits before they land
- The Zeroshot website says the project is starred by engineers at Google, Meta, Shopify, Uber, Booking.com, GitHub, Cloudflare, Atlassian, and JetBrains; this is evidence of community interest rather than a disclosed customer list.