Watch all the sessions on YouTube

Nimrod Kor Nimrod Hauser

Building a Retrievable Codebase Memory Model

A Talk by Nimrod Kor and Nimrod Hauser

Proudly supported by

About this Talk

Accurately modeling module relationships in a codebase is crucial for transitioning manual development tasks into agentic workflows. This session presents research on leveraging graph-based representations of source code to create structured, persistent memory of codebases for developer-agents.

We will examine the value of systematic extraction and parsing of code relationships using Abstract Syntax Trees (ASTs) for extracting relational metadata, as-well as pandas-based data aggregation methods for managing large-scale relational data. Further, we discuss how graph theory techniques including centrality measures and clustering (implemented with NetworkX) can be employed to identify critical software modules and dependencies.

This session references and builds upon previous research, including RepoAgent (arXiv:2402.16667), ContextModule (arXiv:2412.08063), and open-source projects such as Potpie and Blarify, highlighting differences, strengths, and areas for improvement.

Last, we'll explore the integration of the memory model with an agentic architecture focusing on fine-tuning and parameter stabilization for consistent module summarization to improve agent-to-agent communication.

  • Learn precise methodologies to systematically extract and represent semantic and syntactic relationships within complex codebases using graph theory and AST-derived approaches.
  • Understand strategies to mitigate variability in model outputs, ensuring consistent results.
  • Practical Applications: Empirical evaluations, empirical evaluations, including reproducibility challenges and solutions.

13 May 2025, 04:45 PM

AI Today

04:45 PM - 05:15 PM

About The Speakers

Nimrod Kor

Nimrod Kor

Co-founder & CTO, Baz

Driving open source and cloud security forward, with deep roots in code analysis and shaping IaC through projects like Checkov and Yor.

Nimrod Kor

Nimrod Hauser

Nimrod Hauser

Principal Software Engineer, Baz

Startup first-hire turned system-building Swiss Army knife, bridging code and data across defense, cyber, crypto, and dev tools.

Nimrod Hauser

Tessl

Tessl’s AI native development platform delivers secure, high-quality, and auto-maintained code — all built to your specs.

Main Sponsor

All Hands AI

Open Source Agents for Developers

Main Sponsor

Anyshift.io

Context-Aware DevOps AI Agent

Main Sponsor

Baz

Purpose-built coding agent for pull requests.

Main Sponsor

Codurance

Codurance is a global software consultancy helping customers achieve faster growth with modern software, people and practices. We cut through the AI hype to deliver tangible, lasting value.

Main Sponsor

Dosu

Dosu gives agents and teams the comprehensive, up-to-date internal knowledge they need.

Main Sponsor

Equal Experts

Equal Experts is a global network of experienced technology consultants specialising in agile software delivery, digital transformation & data-driven solutions. Known for its collaborative, high-expertise approach, they help organisations solve complex challenges and build lasting capabilities.

Main Sponsor

Qodo

Qodo is an agentic code integrity platform for reviewing, testing, and writing code, integrating AI across development workflows to strengthen code quality at every stage.

Main Sponsor

RunWhen

RunWhen syncs your K8s or cloud account to a massive registry of automated tasks in 15 minutes. Seamlessly connect these tasks to your alerts, tickets, and chats to create real time workflows, automate responses, and intervene before escalating to a human.

Main Sponsor

Stakpak

Specialized DevOps intelligence with a web IDE and Agents

Main Sponsor

Want to sponsor this event? Contact Us