OpenBI
AI-Native Business Intelligence Platform
Start new thread

OpenBI - AI-Native Business Intelligence Platform

by•
OpenBI is a self-hosted, open-source AI-native BI platform. Connect 90+ data sources (Postgres,MySQL & more), then ask questions in plain English - no SQL needed. What makes it different: • AI agents with text-to-SQL + RAG over your documents • Auto-generated AntV G2 charts from chat results • Drag-and-drop dashboards with real-time WebSocket updates • PDF & AI-powered PPTX export (Presenton) • Cron-scheduled reports via Email, Telegram & Webhook • Full Telegram bot for mobile access

Add a comment

Replies

Best
Maker
📌
Hey Product Hunt! 👋 I'm Narender, and I built OpenBI out of frustration - every open-source BI tool I tried either required writing SQL, setting up a data team, or sending data to a third-party cloud. I wanted something truly self-hosted where anyone on the team could just ask a question and get an answer. Here's what OpenBI does differently: 🤖 AI Chat with your data — type any question in plain English. MindsDB translates it to SQL, runs it against your live database, and streams back the answer token-by-token with the exact SQL shown so you always know what ran. 📊 Auto-generated dashboards — chat results automatically become AntV G2 charts or pivot tables. Drag, resize, and rearrange widgets on a Gridstack grid. Every widget update broadcasts live over WebSocket so your whole team sees changes in real time. 🧠 Knowledge Base RAG — upload PDFs, CSVs, Excel, Markdown docs into a vector store. Your AI agents query structured databases AND unstructured documents in the same response. Great for combining live sales data with board reports or policy documents. 📄 PDF & PPTX export — one-click PDF capture stored in MongoDB GridFS with full version history. PPTX export uses Presenton (self-hosted AI slide generator) — it reads your dashboard context and generates editable slides in seconds. 📅 Scheduled reports — set a cron expression, pick a dashboard, choose Email + Telegram + Webhook delivery. Reports run automatically, refresh all widget data, and deliver a PDF attachment. No Celery, no Redis queue — just APScheduler running inside FastAPI. 📱 Telegram bot — full mobile interface. Browse dashboards, ask questions, receive PDF/PPTX files, switch projects and agents — all from Telegram. 🔌 90+ data sources — Postgres, MySQL, Snowflake, BigQuery, Stripe, HubSpot, Shopify, GitHub, Notion, and many more. Credentials are encrypted at rest with Fernet — nothing stored in plain text. 🛡️ Fully self-hosted — runs in Docker Compose with 6 services. Your data never leaves your infrastructure. MIT-style open source (CC BY-NC-ND). I'd love feedback on: → The multi-agent routing (mention multiple agents for parallel queries) → The AI PPTX export quality via Presenton → Any data source you wish was supported Drop your questions below — I'm here all day! 🚀

   Really interesting combination of AI-native analytics and self-hosting. The RAG layer over documents alongside text-to-SQL stands out, especially for teams whose knowledge is split between databases and internal docs. How do you handle accuracy and validation when an answer pulls from both structured data and unstructured documents in the same query?

Maker

   

That's a great question - validating hybrid queries is a complex topic. In OpenBI, we use a multi‑layer architecture that combines classic RAG testing principles with our own application‑specific checks.

Our approach uses a layered evaluation strategy for the different stages of a hybrid query: Data Ingestion, Retrieval, Generation, and Post‑Generation.

Layer 1: Data Ingestion Validation

Accuracy starts with the data itself.

* Structured Data (SQL): We use MindsDB's native text‑to‑SQL capabilities, which are validated on internal benchmarks.

* Unstructured Data (RAG): Documents are processed using a summarization module that produces a compact but information‑dense natural language representation, providing grounding context for all future operations.

Layer 2: Retrieval Validation

Hybrid queries require the system to know where to look.

* Structured Schema Retrieval (Text‑to‑SQL): We employ a schema‑guided RAG approach, combining structured knowledge graphs for relational reasoning with classic RAG for contextual grounding. This improves accuracy and explainability.

* Document Retrieval (Document RAG): We use standard `precision` and `recall` metrics to evaluate the quality of retrieved text chunks.

Layer 3: Generation Validation

We evaluate the final output for both syntactic and semantic correctness.

* Syntactic Validity: SQL generation is verified using `sqlparse`, acting as the first gate against malformed queries.

* Semantic Correctness and Intent Alignment: Both are evaluated using an LLM‑based assessment to check the logic and ensure the output truly addresses the user's original question.

Layer 4: Post‑Generation Validation & Observability

The final step is a final verification, supported by optional LLM call tracing via Langfuse or Normal DB storage. This provides observability across the entire system, allowing for performance audits and debugging.

Ultimately, it's this combination of techniques—from schema‑guided retrieval and modular evaluation to end‑to‑end LLM tracing -that allows OpenBI to deliver reliable answers from hybrid queries, even when the answer relies on disparate sources.

To have a good system we can add the LLM as Judge on top of each query - but ideally its on business requirement and use cases