Julep AI

The next YC application is due in 6 days. We applied to YC 5 times. The first 4 were with Julep. We got 2 interviews, had revenue, and at one point were close to $400K ARR. We still didn’t get in. Since we announced we're doing YC with Memory Store this batch, a lot of founders reached out asking what changed this time. So I wrote about it: - why ARR was not enough - how we treated the application like a product - why user insights mattered most - what I’d tell repeat applicants The biggest lesson for me: the application got better when the company got sharper. Not because we found the perfect wording, but because we had clearer answers to who we were building for, what we had learned, why Memory Store needed to exist, and why now. I hope this is useful for founders applying this batch. P.S. If you’re a female founder applying this batch and want another pair of eyes on your application, please reach out. I’d be happy to help where I can.

Soo proud of the team - we're in Y Combinator!!!! When building memory for julep, we realized memory as a dev tool for each agent is not the bottle neck. A shared memory for teams, that can be then fed to any agent is what's needed. That's what the team is building with Memory Store (YC P26)!!! Julep is still fully open source.

We've been working on something new! Your team's real-time conversations hold more institutional context than any wiki or document. Decisions get made in Slack threads. Architecture gets debated in channels. Onboarding context lives in customer channels. All of that knowledge is ephemeral. It fades into infinite scroll, unreachable by search, invisible to your agents. We've been working on shared team memory that connects directly to where your team already communicates. Daily sync. Cross-tool context. Your agents sharing the same memory as your team. Zero manual effort. Always up to date. Follow Memory store for more updates.

🔌 Julep Now Supports Model Context Protocol (MCP)! Hey Everyone Exciting news! Julep is now an MCP client, connecting you to the entire ecosystem of MCP-compatible tools and servers! 🚀 What This Means Your Julep agents can now dynamically connect to any MCP server and instantly access its capabilities. No hardcoded integrations needed - just point your agent at an MCP server and all its tools become available automatically. Key Features - Connect to any MCP server - Discover available tools at runtime - Switch between servers without code changes - HTTP Transport (simple request-response) - SSE Transport (real-time updates & long-running ops) - Works with Julep's auto_run_tools for autonomous agents - Normalized response format across all MCP servers - Full authentication support Here are some use cases our community has already built - agent that interact with GitHub, manage pull requests, review code, and deploy applications - Connect to database MCP servers (Postgres, MySQL) to query data - Use Puppeteer MCP servers to scrape data What You Can Do - Connect to hundreds of existing MCP servers: - GitHub, Slack, Google Drive integrations - Database tools (Postgres, MySQL) - Developer tools (Puppeteer, Git) - Custom MCP servers you build 📖 Get Started Check out the complete guide with examples from the link in the comments Happy building! 🛠️ — The Julep Team

Introducing the Google Sheets Integration with Julep! Our new integration with Google Sheets brings together smart AI agents with the world's most popular collaborative spreadsheet platform. Here's why we're pumped about this: ✅ Seamless data operations (read, write, append, batch update across multiple ranges) ✅ Real-time workflow automation (trigger actions based on spreadsheet changes) ✅ Flexible authentication (use your own service account or ours for quick setup) ✅ Zero code configuration (set up powerful data workflows through simple YAML) Real-world impact: Julep agents can now manage your spreadsheet workflows automatically. Whether you're tracking metrics, managing inventories, processing form submissions, or building dynamic reports - it handles data collection, transformation, and updates without manual intervention. You can dive deeper into the integration and how to leverage it with Julep

Introducing the Unstructured Integration with Julep! Our new integration with Unstructured.io brings together smart AI agents with enterprise-grade document processing capabilities. Here's why we're pumped about this: ✅ Universal document processing (25+ file types from PDFs to audio files) ✅ Intelligent content extraction (semantic chunking based on document structure) ✅ Production-ready infrastructure (dedicated cloud infrastructure, no dependency management) ✅ LLM-optimized output (structured data ready for RAG and AI workflows) Real-world impact: Julep agents can now transform any unstructured document into AI-ready data automatically. Whether you're building knowledge bases from enterprise documents, processing customer feedback at scale, or creating intelligent document workflows - it handles the entire extraction and transformation pipeline seamlessly. You can dive deeper into the integration and how to leverage it with Julep here: https://lnkd.in/gwzrsDdk

Introducing the Algolia Integration with Julep! We've been working on something exciting - our new integration with Algolia brings together smart AI agents with lightning-fast search and discovery capabilities. Here's why we're pumped about this: ✅ Sub-100ms search performance (real-time results from the first keystroke) ✅ AI-powered personalization (tailored results for every user interaction) ✅ Smart recommendations engine (frequently bought together, trending items, related content) ✅ Enterprise-grade scalability (handles billions of queries with global CDN) Real-world impact: Julep agents can now power sophisticated search and discovery experiences automatically. Whether you're building e-commerce product discovery, content recommendation systems, or intelligent knowledge bases - it delivers personalized, superfast results that adapt to user behavior in real-time. You can dive deeper into the integration and how to leverage it with Julep here: https://lnkd.in/g9WdHvvr

Introducing the Brave Search API Integration with Julep! Our new integration with Brave Search API brings together smart AI agents with privacy-first, independent web search capabilities. Here's why we're pumped about this: ✅ Independent web index (real alternative to Google/Bing dependency) ✅ Privacy-first search (no tracking or profiling of your users) ✅ Real-time web data access (fresh results from 8+ billion queries) ✅ Advanced search customization (Goggles for custom ranking and filtering) Real world impact: Julep agents can now access the entire web intelligently while respecting user privacy. Whether you're building research assistants, competitive analysis tools, or AI-powered content discovery, it delivers comprehensive search results without compromising on privacy or independence. You can dive deeper into the integration and how to leverage it with Julep here: https://lnkd.in/gXXhFpWj

Introducing the Cloudinary Integration with Julep! We've been working on something exciting - our new integration with Cloudinary brings together smart AI agents with powerful visual media management capabilities. Here's why we're pumped about this: ✅ Smart image and video transformations (resize, crop, optimize on demand) ✅ AI-powered media optimization (automatic format selection for every device) ✅ Seamless CDN delivery (global performance without the infrastructure headache) ✅ Complete media lifecycle management (upload, store, transform, deliver - all automated) Real-world impact: Julep agents can now handle your entire visual media workflow automatically. Whether you're building e-commerce product galleries, social media content pipelines, or user-generated content platforms, it automatically optimizes and delivers the perfect media experience for every user. You can dive deeper into the integration and how to leverage it with Julep here: https://lnkd.in/gMq-iv_E