Psychic - An open-source integration platform for unstructured data
My cofounder and I used to work at Robinhood where we shipped the company’s first OAuth integrations, so we know a lot about how data moves between companies. For example, we know that the pain of building new API integrations scales with the level of fragmentation and number of competing "standards". In the current meta, we see this pain with a lot of AI startups who invariably need to connect to their customers data, but have to support 50+ integrations before they even scale to 50+ customers. This is the process for an AI startup to add a new integration for a customer: - Pore over the…
In plain words
Psychic is an open-source integration platform that helps AI startups connect to unstructured data across multiple sources. It simplifies the process of building and managing API integrations with various applications, reducing the engineering effort required to support customer data connections. The platform handles API documentation navigation, OAuth setup, and data normalization, targeting AI companies that need to quickly integrate with numerous data sources without building custom connectors for each one.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
My cofounder and I used to work at Robinhood where we shipped the company’s first OAuth integrations, so we know a lot about how data moves between companies. For example, we know that the pain of building new API integrations scales with the level of fragmentation and number of competing "standards". In the current meta, we see this pain with a lot of AI startups who invariably need to connect to their customers data, but have to support 50+ integrations before they even scale to 50+ customers. This is the process for an AI startup to add a new integration for a customer: - Pore over the API docs for each source application and write a connector for each - Play email tag to find the right stakeholders and get them to share sensitive API keys, or give them an OAuth app. It can take 6+ weeks for some platforms to review new OAuth apps - Normalize data that arrives in a different formats from each source (HTML, XML, text dumps, 3 different flavors of markdown, JSON, etc) - Figure out what data should be vectorized, what should be stored as SQL, and what should be discarded - Detect when data has been updated and synchronize it - Monitor when pipelines break so data doesn’t go stale This is a LOT of work for something that doesn’t move the needle on product quality. That’s why we built Psychic.dev to be the fastest and most secure way for startups to connect to their customer’s data. You integrate once with our universal APIs and get N integrations with CRMs, knowledge bases, ticketing systems and more with no incremental engineering effort. We abstract away the quirks of each data source into Document and Conversation data models, and try to find a good balance to allow for deep integrations while maintaining broad utility. Since it’s open source, we encourage founders to fork and extend our data models to fit their needs as they evolve, even if it means migrating off our paid version. To see an example in action, check out our demo repo here: https://github.com/psychic-api/psychic-langchain-tutorial/ We are also open source and open to contributions, learn more at docs.psychic.dev or by emailing us at [email protected]!
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 16d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
Launched alongside, May 2023
the whole month →
- BR
In today's world, catchy headlines and articles often distract readers from the facts and relevant information. By utilizing OpenAI's language models, Boring Report processes sensationalist news articles, transforms them into the content you see, and helps readers focus on the essential details. We recently updated our iOS app experience, so any and all feedback would be appreciated. App Link: https://apps.apple.com/us/app/boring-report-news-by-ai/id644...
AI · 2023 · boringreport.org
Momentum Page▲834Launch your website in seconds, get users in minutes
Dev tools · 2023 · page.mmntm.build