Click here to get this post in PDF
Potpie is building a foundational context layer that allows AI agents to operate across complex, large-scale codebases the way experienced engineers do.
San Francisco, California – February 23, 2026; Software teams are moving faster than ever, yet the systems they build and maintain were never designed for AI agents to operate inside them. Codebases span millions of lines, context is scattered across dozens of tools, and critical knowledge lives in the heads of a few senior engineers. Potpie was built to change that. Today, the company announced a $2.2 million pre-seed round to help engineering teams unify context across their entire stack and make AI agents genuinely useful in complex software environments.
The round was led by Emergent Ventures with participation from All In Capital, DeVC and Point One Capital. The capital will be used to support early enterprise deployments, expand the engineering team, and continue building Potpie’s core context and agent infrastructure.
As generative AI adoption accelerates, most tools focus on surface-level code generation while ignoring the deeper problem of context. Large language models are powerful, but without access to system-level understanding, tooling history, and architectural intent, they struggle in real production environments. Traditional approaches rely on senior engineers to manually hold this context together, a model that breaks down at scale and fails entirely when AI agents are introduced.
Potpie addresses this by unifying context across the entire engineering stack and enabling spec driven development. It pulls in information from source code, tickets, logs, documentation, and reviews, links it together, and makes it usable by agents.
With Potpie, the spec becomes the source of truth, not the existing code. Agents plan the feature end to end first by turning requirements into a clear implementation plan, mapping dependencies and edge cases, and aligning tests and rollout steps before writing a single line of code. The principle is simple. An agent is only as effective as the information it can access and the tools it can use. Potpie focuses on both.
“As AI makes code generation easier, the real challenge shifts to reasoning across massive, interconnected systems. Potpie is our answer to that shift, an ontology-first layer that helps enterprises truly understand and manage their software” said Aditi Kothari, CEO and co-founder of Potpie.“
The platform enables teams to automate high-impact and non trivial use cases across the software development lifecycle, like debugging cross-service failures, maintening and writing end-to-end test, blast radius detection and system design. It is designed for enterprise companies with large and complex codebases, starting at around one million lines of code and scaling to hundreds of millions. Rather than acting as another coding assistant, Potpie builds a graphical representation of software systems, infers behavior and patterns across modules, and creates structured artifacts that allow agents to operate consistently and safely.
Potpie also actively creates context as systems evolve. When pull requests are created, it can update documentation and tickets automatically. When tickets are opened, it can generate system designs. When releases ship, it can produce release notes. The platform generates Agent.md files that define how agents should behave within a codebase and builds a searchable, tagged index across APIs, services, databases, and components to reduce search space and improve reliability.
The company was founded by Aditi Kothari and Dhiren Mathur, who began working on the problem in October 2023, at the start of the first wave of generative AI adoption. While much of the industry focused on knowledge workers, they saw that developers faced a fundamentally different challenge. Code is non-linear, deeply interconnected, and spread across large systems. They spent nearly two years building the foundational layer that understands codebases and creates the underlying knowledge graph, before launching Potpie publicly last year in January 2025
Early deployments reflect the scale of the problem Potpie is addressing. One customer with a codebase exceeding 40 million lines reduced root cause analysis for production issues from nearly a week to around 30 minutes, with engineers acting as reviewers instead of investigators. Another customer maintaining decades-old, hardware-integrated systems used Potpie to update and generate tests in the background, compressing work that previously took multiple sprints into a much shorter cycle.
In large enterprises, the real challenge is not generating code, it is understanding the system deeply enough to change it safely. Potpie’s ontology-first architecture, combined with rigorous context curation and spec-driven development, creates a structured model of the entire engineering ecosystem. This allows AI agents to reason across services, dependencies, tickets, and production signals with the clarity of a senior engineer. That is what makes Potpie uniquely capable of solving complex RCA, impact analysis, and high-risk feature work even in codebases exceeding 50 million lines.
Anupam Rastogi, Managing Partner at Emergent Ventures, commented: In large enterprises, the real challenge is not generating code, it is understanding the system deeply enough to change it safely. Potpie’s ontology-first architecture, combined with rigorous context curation and spec-driven development, creates a structured model of the entire engineering ecosystem. This allows AI agents to reason across services, dependencies, tickets, and production signals with the clarity of a senior engineer. That is what makes Potpie uniquely capable of solving complex RCA, impact analysis, and high-risk feature work even in codebases exceeding 50 million lines.”
Potpie currently works with Fortune 500 and publicly listed companies in regulated industries, including healthcare and insurtech. Its open-source projects have surpassed 5,000 stars on GitHub, creating a strong magnet for enterprise adoption.
“AI readiness is not about picking the right model,” Aditi Kothari added. “It’s about building systems that can support intelligence over time. Our goal is to make Potpie the foundational layer engineering teams rely on to build, operate, and evolve complex software with AI built in from the start.”
About Potpie
Potpie is a foundational context layer that allows AI agents to operate across complex, large-scale codebases the way experienced engineers do.
Potpie pulls in information from source code, tickets, logs, documentation, and reviews, links it together, and makes it usable by agents. In doing so, Potpie is unifying context across the entire engineering stack and enabling spec driven development.
With Potpie, the spec becomes the source of truth, not the existing code. Agents plan the feature end to end first by turning requirements into a clear implementation plan, mapping dependencies and edge cases, and aligning tests and rollout steps before writing a single line of code. The principle is simple. An agent is only as effective as the information it can access and the tools it can use. Potpie focuses on both.
Also read:
AirConsole Launches with Škoda, Introducing a Škoda-Exclusive In-Car Tetris® Experience
Image source: Potpie.ai
