SearchHack: Build the next layer of AI search
Welcome to SearchHack: Build the Next Layer of AI Search
AI is changing how we search—but the next breakthrough will not come from simply building a better search box.
SearchHack brings together builders, researchers, engineers, and entrepreneurs to rethink how information is discovered, retrieved, routed, remembered, and reasoned over in the age of AI.
Today’s AI systems face a new class of search problems: too much information, fragmented context, expensive retrieval, weak personalization, limited memory, and poor routing across increasingly complex sources. Traditional keyword search is no longer enough—and simply adding an LLM on top of search is not the final answer.
Judges - TBD
Prizes
First Place: iPhone 18 Pro + 30,000 Tavily credits
Second Place: Apple Watch Series 12 + 20,000 Tavily credits
Third Place: AirPods 5 + 10,000 Tavily credits
⏱️ 6-Hour Schedule
12:00–12:15 — 🚀 Kickoff Challenge introduction, rules, judging criteria & team formation
12:15–4:15 — 🛠️ Build — 4 Hours Build your prototype with mentors available throughout the session
4:15–4:30 — 📦 Submission & Demo Prep Finalize your project and prepare your demo
4:30–5:00 — ⚖️ Judging — 30 Min Judges evaluate projects across innovation, technical execution, search intelligence, and impact
5:00–6:00 — 🎤 Final Presentations & Winner Announcement
The Challenge
Build the next layer of AI search. Your solution could rethink any part of the search stack—from query understanding and intelligent routing to long-context retrieval, agentic search, memory, ranking, multimodal discovery, knowledge graphs, or entirely new interfaces for interacting with information. -What if search could understand what you mean not just what you type?
What if it knew where to look, what to remember, what to ignore, and when to keep searching?
What if search became an intelligent system rather than a list of links?
We encourage teams to explore ideas such as:
🔎 Intelligent Search & Retrieval — Go beyond keyword matching.
🧠 Long-Context & Memory — Give AI the ability to find and use information across massive context.
🧭 Search Routing — Decide which search engine, database, tool, or source should answer a query.
🤖 Agentic Search — Let AI plan, search, verify, and synthesize across multiple sources.
⚡ Efficient Search — Reduce latency, token consumption, and unnecessary retrieval.
🕸️ Knowledge & Discovery — Connect fragmented information into meaningful relationships.
🎯 Personalized Search — Understand the user's context, intent, and history.
👁️ Multimodal Search — Search across text, images, video, audio, code, and other data.
🚀 New Search Interfaces — Reimagine what it means to interact with information.
The Goal
Build the infrastructure, intelligence, or experience that makes AI search fundamentally better. Whether you're improving an existing search stack or creating something entirely new, SearchHack is a space to experiment with the technologies that could define the next generation of information discovery. Search is no longer just about finding information. It's about knowing what to find, where to find it, and what to do with it. Come build the next layer.