Skip to main content
Back to Projects
LLMSep 2024 → Present

NL2SQL & Data Intelligence Chatbots

Chatbots that turn natural language into SQL, Pandas, and Neo4j Cypher queries against structured and relational data, with an Excel-AI-Chatbot on top of DuckDB.

PythonDuckDBPandasNeo4jFastAPILangChainopenpyxl
Sources
DuckDB · Neo4j · Excel
Query accuracy
92%
Avg response
1.8s

Problem & Solution

The Problem

Business users want to ask questions of their data in plain language, but the data lives in DuckDB warehouses, Excel files, and Neo4j graphs — and they don't know how to write SQL, Pandas, or Cypher.

The Solution

Built a chat surface that classifies the data source, generates the right query (SQL, Pandas, or Cypher), runs it against DuckDB / Neo4j, and returns a plain-language answer. The Excel-AI-Chatbot ingests spreadsheets into DuckDB on the fly for analysis.

System Architecture

End-to-end flow from intake to outcome

  1. 01

    Source Classifier

    LangChain

    LLM classifies the user's question into SQL, Pandas, or Cypher intent.

  2. 02

    DuckDB Engine

    DuckDBPandas

    SQL or Pandas generated and executed against DuckDB.

  3. 03

    Neo4j Engine

    Neo4j

    Cypher generated and executed against Neo4j for graph queries.

  4. 04

    Excel Ingest

    openpyxlDuckDB

    Excel files are ingested into DuckDB for analysis on the fly.

Engineering Challenges & Wins

Optimization results and engineering tradeoffs

Optimization Results

  • Routes between SQL, Pandas, and Cypher based on the user's question.
  • Excel-AI-Chatbot ingests spreadsheets into DuckDB on the fly.
  • Neo4j integration for relational graph queries.

Engineering Challenges

  • Choosing the right query language for ambiguous natural-language questions.
  • Keeping query latency low for large DuckDB datasets.

More case studies

Explore other production systems I've engineered end-to-end.

All Projects

Related Projects

LLM
AI Avatar Assistant (Real-Time)

Real-time avatar-based assistant combining LiveKit, WebRTC, Groq inference, and TTS/STT, with a Ready Player Me avatar for natural face-to-face interactions.

Streaming latency
<300ms
Concurrent sessions
200+
Avatars supported
RPM SDK
PythonFastAPILiveKitWebRTCGroq+3
View Case Study
Autonomous AI AgentsFeatured
OmniShield ICES: Asynchronous AI Email Security Middleware

Mail-server gateway (Postfix Milter) that intercepts every inbound message, runs a fine-tuned BERT core (M-BSCE) through an asynchronous inference pipeline (RabbitMQ + Celery), and adds multimodal threat detection (PaddleOCR for Quishing) with SOAP clawbacks to the upstream mail system.

Quishing detection
98%+
p95 classification
<400ms
Async throughput
5k msg/min
PythonPostfix MilterRabbitMQCeleryBERT+6
View Case Study