Advanced RAG Knowledge Assistant
A document question-answering assistant with semantic chunking, reranking, query rewriting and a validated FastAPI layer.
- Python
- LangChain
- ChromaDB
- Embeddings
AI Lab
Each project shows whether a working deployment is connected. Tools marked “runs in browser” work right here with no server or API key.
In-browser tool
See how chunk size and overlap change the passages a RAG retriever would index — the same preprocessing step used in the Knowledge Assistant.
87 words
Retrieval-augmented generation pairs a language model with a search step. Documents are split into chunks, each chunk is embedded as a vector, and the
embedded as a vector, and the vectors are stored in a database such as ChromaDB. When a question arrives, the system embeds the question,
arrives, the system embeds the question, retrieves the most similar chunks, and passes them to the model as context. Chunk size matters: chunks that
context. Chunk size matters: chunks that are too small lose meaning, while chunks that are too large dilute the signal. Overlap keeps sentences that
the signal. Overlap keeps sentences that sit on a boundary from being cut in half.
A document question-answering assistant with semantic chunking, reranking, query rewriting and a validated FastAPI layer.
An assistant that combines text, voice and images, orchestrating tool calls against a SQLite-backed flight workflow.
Converts Python to C++ or Rust through a provider-independent interface, then compiles and runs the output to compare models.
A structured workflow for generating SQL with ChatGPT, then validating every query against business logic in MySQL.