Viktor.comAn AI coworker that actually does the work
Promoted
Phaide AI
Stop building static dashboards. Start discovering.
18 followers
Stop building static dashboards. Start discovering.
18 followers
Connect 750+ data warehouses, CRMs, and business tools. Get scheduled agentic reports and real-time dashboards keep your team on top of every KPI.
Hi Product Hunt 👋 Shintaro here, founder of Phaide AI.
Before this, our team built quelmap — an open-source, fully local data analysis agent — and Lightning-4b, a 4B-parameter LLM fine-tuned for analysis. Talking to the people who used them, the same two problems kept coming up:
1. Most teams don't know what to ask their data. The insight they needed was usually sitting right there — behind a question nobody thought to ask.
2. Many companies can't hand confidential data to an AI at all, so "AI analysis" was a non-starter before it began.
Phaide AI is our answer to both.
What's live today at phaide.ai:
- Connect CSV/Excel, Postgres, MySQL, Supabase, BigQuery, Snowflake, and more — and analyze across sources in one place.
- Autonomous exploration: a hypothesis-validation tree where each node is an independent agent with its own hypothesis and its own Python sandbox. It tests the idea, then branches. Findings become a report with charts — plus a separate pass that extracts the genuinely surprising discoveries.
- Automatic masking at ingest: confidential values are hashed to anonymized tokens before any model sees them. The mapping never leaves your org.
- Dashboards built by the agent and edited via chat, PowerPoint/Excel/Word export, scheduled explorations delivered to Slack or email.
- A developer surface: REST API and an MCP server, so your own tools and agents can drive Phaide.
Free tier is 3 users, 400 ai credits, no credit card — you can try it on your own data in a few minutes.
A question for you: what's a problem in your data that you found too late — something you wish an agent had flagged before you knew to look? That's exactly the kind of case we want to throw at the exploration tree, and I'll be in the comments all day.
Hi Product Hunt 👋
My focus has been turning Phaide’s autonomous analysis into something teams can trust with real, messy, and confidential business data—not just polished demo datasets.
We spent a lot of time building the infrastructure behind the experience: secure data connections, isolated Python environments for every analysis agent, automatic data masking, and a system that lets multiple agents test different hypotheses in parallel.
The part I’m most excited about is that Phaide doesn’t simply wait for the perfect prompt. It explores the data, challenges its own hypotheses, and surfaces findings you may never have thought to search for.
We’re still early, so your feedback would mean a lot. If you try Phaide today, please tell us where it surprised you—and where it fell short. I’ll be here answering technical questions and learning from everyone’s use cases throughout the launch! 🚀