The goal is to build cohesive or contextual smart agents leveraging LLM.
Prerequisite
- install dependencices
pip install -r requirements.txt
- create a .env file with
OPENAI_API_KEY=<your key>
KAI is mostly an LLM over set of documents (pdf, doc, html, epub etc). It leverages Embedding-based search to answer question from the context (documents).
TODO: Test other file formats for context e.g. ppt, doc, html, epub etc
Add some pdf's for context in ./docs folder
reference: OpenAI cookbook
python kai.py clean --all
BAAI can help breaking down business features into Epics > Stoires > Acceptance cirterias.
TODO: extend the solution to groom stories with more validations and edge scenerios
baai_context.br.py have top 5 features on building Singapore Rental car system extarcted from GPT. The questions that were asked to GPT were (1) what are the probelms with Singapore rental car system (2) Given above context what are the top 5 features to implement
reference: deeplearning.ai’s ChatGPT Prompt Engineering for Developers course
python baai.py clean --all
This is just an alternate approach to use FastAPI to expose the application as REST endpoint
python main.py clean --all
Pulse Check
curl -X GET http://localhost:8001/pulse
Ask Question
curl -X POST -H "Content-Type: application/json" -d '{"source": "unknown", "question": "What is boo?"}' http://localhost:8001/ask