Factory builds the Command Center for software development with OpenAI’s reasoning models

Factory⁠(opens in a new window), founded in 2023 by Matan Grinberg and Eno Reyes, uses OpenAI’s reasoning models to change how organizations build and maintain software.

The Factory Platform is designed to eliminate the bottlenecks in software development. With OpenAI o1, o3‑mini, and GPT‑4o embedded directly into the Platform’s workflows, Factory is building a development environment that goes beyond writing code to understanding, retrieving, and reasoning through complex systems, achieving:

• 2–4x faster feature development cycles.
• 60% reduction in context switching time.
• 10+ additional hours per week per developer across the software development lifecycle.

“The flexibility of offering o1 and o3‑mini for a range of reasoning tasks has helped us win business from customers seeking a software development tool that allows seamless switching between models with varying reasoning capabilities,” says Eno Reyes, Factory’s co-founder and CTO.

Software development has often been constrained by manual research, fragmented knowledge, and slow iteration cycles. Engineers don’t just write code; they spend hours understanding dependencies and verifying system accuracy.

Before OpenAI’s family of reasoning models, Factory relied on a mix of in-house solutions and traditional tooling, but they ran into key limitations:

• Planning bottlenecks: No automated way to structure and coordinate development steps.
• Code editing inefficiencies: Existing AI completions lacked the speed and accuracy needed for large-scale development.
• Ineffective knowledge retrieval: Standard search tools didn’t surface the right code snippets or documentation efficiently.

Factory saw an opportunity to change this by building a system where AI goes beyond suggesting code to understanding and reasoning about the entire software development process. They work closely with OpenAI to integrate cutting-edge models at every stage of the software development lifecycle.

Factory has integrated OpenAI's models strategically across different stages of the software development lifecycle, selecting specific models based on their reasoning capabilities, speed, and accuracy requirements for each task:

“We've found that giving developers flexibility to choose the right model for their specific task is crucial,” says Reyes. “For quick code reviews, o3‑mini provides almost identical quality to larger models but at significantly faster speeds, while complex architectural planning benefits from o1's deeper reasoning capabilities.”

Factory is also experimenting with reinforcement fine-tuning o3‑mini for code reranking and auto-injecting lightweight guidance to improve model compliance—enhancing precision in production-grade AI development.

With OpenAI, Factory doesn’t just write code—it reasons through engineering problems, retrieves critical knowledge, and optimizes workflows.

Factory’s context-first architecture dynamically pulls insights from codebases, documentation, and issue tracking systems, minimizing context-switching and cognitive overhead for developers.

Instead of switching between multiple tools, developers can find everything they need in one place, allowing them to focus on high-leverage work.

Factory's next focus is making AI-powered development even more autonomous.

Through integrating native tools across source control, project management, team communication, error monitoring, and continuous delivery pipelines, combined with reasoning models, they're building AI systems that can proactively plan, execute, and refine engineering tasks across the entire software development lifecycle.