Polyvia
Queryable visual knowledge index for agents
94 followers
Queryable visual knowledge index for agents
94 followers
Polyvia is the first Visual Knowledge Index for Agents & MCPs. Turn scattered visuals into a queryable source of truth with every fact disambiguated. Other tools extract visuals OR index text — Polyvia indexes and reasons over visuals, connecting facts across 10,000s of documents. Built for developers of multimodal agents and knowledge-work teams.
Polyvia
Hey Product Hunt 👋, this is Mateusz from the Polyvia AI team, and we're very excited to launch Polyvia today!
🚀 Get early access → polyvia.ai/#access (or send email to: mateusz@polyvia.ai)
The problem: ~80% of enterprise knowledge lives in visual form — charts, slides, diagrams, infographics. Current AI tools either extract visuals OR index text - neither can reason across thousands of visual documents.
The result? Agents hallucinate. Search fails. Insights stay buried.
The solution: Polyvia is the first Visual Knowledge Index for Agents & MCPs - queryable index of facts for visual data. Think "Pinecone for visual data" - but with agentic visual reasoning.
Turn scattered visuals into a queryable source of truth.
---
What Polyvia does:
• Extraction — Charts, tables, diagrams, infographics → structured visual logic
• Indexing — Connect facts into a graph-based single source of truth
• Reasoning — Query for answers with audit-ready citations, or explore topics
Example:
Query: "What's the difference between revenue of Company-A and Company-B in the last quarter?"
❌ Text-Only Agent: "I cannot find specific revenue numbers in the text."
✅ Polyvia:
→ "Company-A Q4 revenue: $10.67M [Chart; page 12; doc 6367]"
→ "Company-B Q4 revenue: $7.05M [Infographic; page 91; doc 65220]"
→ "Difference: $3.62M"
2 ways to access:
• 🔌 Polyvia API & Polyvia MCP Server — available through REST API and our MCP Server for Claude, Cursor, and more
• 🖥️ Polyvia Studio — visual search and exploration for knowledge-work teams in research, finance, healthcare
---
We are currently in private Beta (Polyvia-0.5). We will be fast-tracking ProductHunt users throughout the day.
🚀 Get early access → polyvia.ai/#access (or send email to: mateusz@polyvia.ai)
📖 Read the full story → Polyvia Blog - Rethinking RAG for Visual Data
The "charts live in PDFs that no RAG can touch" problem is very real. We've had situations where the answer to a question was clearly visible in a chart, but the AI kept saying "I cannot find that information."
The MCP integration is smart - means you can plug this into existing Claude workflows without rebuilding everything. Curious about the latency - when you're querying across 10,000+ documents with visual reasoning, how fast is the response? Is this more of a batch/research use case or can it handle real-time agent workflows?
Polyvia
@philip_sorensen
It's built for real-time agent workflows, not just batch research.
On moderately-sized datasets, both ingest and query are fast enough for production use.
For 10k+ docs, the index is pre-built so queries stay snappy — you're not re-processing visuals on every call.
The MCP integration means you can test this in your existing Claude setup in a couple minutes.
Happy to help you benchmark on your own docs if you want to try it.
-- Mateusz
CreateOS
Congrats on the launch! I just went through the site and watched the demo, the way you guys handle reasoning over visuals across thousands of documents is seriously impressive. Integrating this via an MCP server for Claude and Cursor is a game-changer for building multimodal agents. Fantastic concept.