No comments yet.
We’re Ghita and Nicholas, the founders of ZeroEntropy.
We are on a mission to build the world’s most accurate search engine over complex and unstructured documents, behind a simple API.
If you are building RAG, a search bar, or search tools for your AI Agents, please reach out to us here founders@zeroentropy.dev
*Why* Most AI products - whether Q&A bots or AI agents - depend on retrieval systems to provide relevant context from knowledge bases.
Yet, the vast majority of these systems rely on basic semantic or hybrid search methods, which still frequently fail.
These mistakes lead to inaccurate responses and hallucinations by LLMs, frustrating developers and end users alike.
That’s why we’re building ZeroEntropy: to add intelligence to retrieval, and empower developers to build reliable and accurate AI products!
*A few common failure modes* The most common examples are:
- Negated Queries, like: “Which electric vehicle articles do not include any reference to Elon Musk?”. - Multi-Hop Queries, like:“Where was the suspect who was identified in the report born?” - Fuzzy Filtering Queries, like: “What diagnostic methods are suggested for early-stage cancer, in papers that mention a sample size of over 2000?”
These types of queries are common in industries like legal, healthcare and manufacturing, where documents are very complex. Semantic similarity and keyword-based approach fail on all these examples.
*Our Solution* Search AI agents can solve these complex queries. Not only can they understand the user’s intent, but also route each query to the most suitable retrieval system, ensuring precise and relevant answers. Moreover, they adapt and personalize their performance over time, continuously improving with use.
All of this capability is accessible through a simple API that you can try out at: dashboard.zeroentropy.dev.
*About us* We both have a background in math and AI: Ghita has two masters in Applied Maths from Ecole Polytechnique (Paris) and from UC Berkeley, and Nicholas grew up doing math and coding competitions at CMU before becoming the CTO of several startups.
We’ve both experienced the pain of building great retrieval systems from working as AI engineers and researchers in several industries from healthcare, to finance, to B2B SaaS, and consumer AI.
*PS/Fun Fact* In case you’re curious about our name, in information retrieval, entropy refers to the expected variability of information contained in a variable or message. ZeroEntropy refers to knowing exactly what the message contains and is associated to a high predictability, low chaos state. In physics, absolute zero entropy is impossible to achieve. But we’ll try!
