This content originally appeared on HackerNoon and was authored by Oleksii Bondar
Welcome to the Proof of Usefulness Hackathon spotlight, curated by HackerNoon’s editors to showcase noteworthy tech solutions to real-world problems. Whether you’re a solopreneur, part of an early-stage startup, or a developer building something that truly matters, the Proof of Usefulness Hackathon is your chance to test your product’s utility, get featured on HackerNoon, and compete for $150k+ in prizes. Submit your project to get started!
Today we are interviewing Oleksii Bondar, the creator behind PMB (Personal Memory Brain). PMB is an innovative, local-first persistent memory solution designed to seamlessly connect to AI coding agents via the Model Context Protocol (MCP). It aims to solve the frustrating bottleneck of developers having to repeatedly explain their codebase to AI agents across different sessions.
What does PMB do? And why is now the time for it to exist?
PMB is local-first persistent memory for AI coding agents, connected over MCP. It gives Claude Code, Cursor, and Codex a shared, durable memory of your project (decisions, facts, corrections, goals) that survives across sessions, using hybrid BM25 + vector + graph retrieval that runs entirely on your machine with no cloud and no API keys. It helps developers stop re-explaining their codebase to the agent every time they open a new session. Now’s a good time for PMB (Personal Memory Brain) to exist because the adoption of AI coding agents is skyrocketing, yet their ephemeral memory between sessions remains a significant bottleneck for developer productivity.
What is your traction to date? How many people does PMB reach?
PMB just hit its 1.0 release (June 2026) and is in early launch, so it is pre-traction on raw monthly numbers. The addressable audience is large and growing fast: developers who use AI coding agents (Claude Code, Cursor, Codex and 6 others) - a segment now in the millions. Distribution is already live via PyPI (pip install pmb-ai) and the official MCP registry, which auto-propagates to the GitHub MCP Registry and other directories, so reach is expected to grow organically through registry discovery and developer word of mouth.
Who does PMB serve? What’s exciting about your users and customers?
Developers who use AI coding agents daily (Claude Code, Cursor, Codex) and are tired of re-explaining their project every session. Especially those who care about privacy and ownership: PMB keeps all memory local (no cloud, no API keys), so it fits people on proprietary or sensitive codebases - indie devs, startups, and teams under compliance constraints. As a newly released open-source project, it has no notable customers yet.
What technologies were used in the making of PMB? And why did you choose the ones most essential to your tech stack?
PMB is built using Python and the Model Context Protocol (MCP), utilizing FastMCP to easily plug into major AI coding agents. To ensure local-first, on-device execution without cloud dependencies, it leverages SQLite for events and entity graphs, LanceDB as a robust vector store, sentence-transformers for MiniLM embeddings, and rank-bm25 for lexical retrieval. This powerful combination enables efficient hybrid BM25, vector, and graph retrieval while keeping all codebase data fully secure and completely under the developer's control.
PMB (Personal Memory Brain) scored a 70 proof of usefulness score(https://proofofusefulness.com/report/pmb-personal-memory-brain)
What excites you about this PMB's potential usefulness?
AI coding agents are everywhere now, but they still forget everything between sessions - you re-explain your project every day. PMB fixes that with memory that is local-first and portable across every agent you use, so your project knowledge is not locked into one vendor's cloud. What excites me most is the ownership angle: as agents get more capable, the memory layer becomes the valuable part, and PMB keeps that layer on the developer's own machine, under their control. Cross-session memory is clearly becoming standard, and a local, open, agent-agnostic option matters.
Meet our sponsors
Bright Data: Bright Data is the leading web data infrastructure company, empowering over 20,000 organizations with ethical, scalable access to real-time public web information. From startups to industry leaders, we deliver the datasets that fuel AI innovation and real-world impact. Ready to unlock the web? Learn more at brightdata.com.
Neo4j: GraphRAG combines retrieval-augmented generation with graph-native context, allowing LLMs to reason over structured relationships instead of just documents. With Neo4j, you can build GraphRAG pipelines that connect your data and surface clearer insights. Learn more.
Storyblok: Storyblok is a headless CMS built for developers who want clean architecture and full control. Structure your content once, connect it anywhere, and keep your front end truly independent. API-first. AI-ready. Framework-agnostic. Future-proof. Start for free.
Algolia: Algolia provides a managed retrieval layer that lets developers quickly build web search and intelligent AI agents. Learn more.
This content originally appeared on HackerNoon and was authored by Oleksii Bondar
Oleksii Bondar | Sciencx (2026-08-18T04:02:07+00:00) PMB (Personal Memory Brain) Earns a 115 Proof of Usefulness Score by Building Local-First Persistent. Retrieved from https://www.scien.cx/2026/08/18/pmb-personal-memory-brain-earns-a-115-proof-of-usefulness-score-by-building-local-first-persistent/
Please log in to upload a file.
There are no updates yet.
Click the Upload button above to add an update.