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:
“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.
## Tackling inefficiencies in software development
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:
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.
## Optimizing the software development lifecycle with the right models
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:
Development stageKey tasksOpenAI modelsWhy this model ExplorationUnderstanding codebases, searching documentation OpenAI o3-mini Fast responses (10x quicker than larger models) with sufficient reasoning for contextual code understanding PrioritizationBug triage, feature analysis OpenAI o3-mini Balance of reasoning capability and speed for evaluating complex dependencies PlanningArchitecture decisions, system design OpenAI o1 High reasoning capabilities for complex architectural decisions and system-level planning ExecutionCode generation, editing, reviews OpenAI o1
GPT-4o Using predicted outputs reduces latency by 50% for real-time coding assistance
“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.
> “OpenAI’s reasoning models allow us to move beyond just code generation. We’re building an agentic development environment where AI can deeply understand, retrieve, and act on engineering knowledge.”
Eno Reyes, Co-founder and CTO of Factory
_Reasoning models like o1 and o3‑mini power codebase understanding and documentation._
## Faster developing and better engineering outcomes with OpenAI
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.
## Building for autonomous software development
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.
> “The software of the future will be built by humans and AI, together, in one platform. With OpenAI’s reasoning models, we’re enabling developers to focus on higher-leverage work while AI handles the complexity.”
Matan Grinberg, Co-founder and CEO of Factory
## Interested in learning more about ChatGPT for business?
Supporting sellers with enhanced product listings API Feb 27, 2025
Building an autonomous financial analyst with o1 and o3-mini API Feb 27, 2025
Using OpenAI o1 for financial analysis API Feb 13, 2025
Our Research * Research Index * Research Overview * Research Residency * OpenAI for Science * Economic Research
Latest Advancements * GPT-5.3 Instant * GPT-5.3-Codex * GPT-5 * Codex
Safety * Safety Approach * Security & Privacy * Trust & Transparency
ChatGPT * Explore ChatGPT(opens in a new window) * Business * Enterprise * Education * Pricing(opens in a new window) * Download(opens in a new window)
Sora * Sora Overview * Features * Pricing * Sora log in(opens in a new window)
API Platform * Platform Overview * Pricing * API log in(opens in a new window) * Documentation(opens in a new window) * Developer Forum(opens in a new window)
For Business * Business Overview * Solutions * Contact Sales
Company * About Us * Our Charter * Foundation * Careers * Brand
Support * Help Center(opens in a new window)
More * News * Stories * Livestreams * Podcast * RSS
Terms & Policies * Terms of Use * Privacy Policy * Other Policies
(opens in a new window)(opens in a new window)(opens in a new window)(opens in a new window)(opens in a new window)(opens in a new window)(opens in a new window)
OpenAI © 2015–2026 Manage Cookies
English United States