Skip to content

Release: Merge develop into main (Docker image size: 5.37GB) - #1

Merged
FairozaAmira merged 18 commits into
mainfrom
develop
Feb 21, 2026
Merged

Release: Merge develop into main (Docker image size: 5.37GB)#1
FairozaAmira merged 18 commits into
mainfrom
develop

Conversation

@FairozaAmira

@FairozaAmira FairozaAmira commented Feb 21, 2026 •
edited
Loading

Copy link
Copy Markdown
Owner

Overview

This is the first pull request from develop into main, so it includes all foundational work completed on the develop branch. Docker file size 5.4GB

Scope

  • Bootstrap FastAPI application structure (controllers, routers, utils, tests)
  • Implement RAG pipeline and retrieval components
  • Add chatbot, crawler, and evaluation API flows
  • Add sample domain datasets and web sample sources
  • Add environment templates for UAT/production
  • Add Docker runtime/build setup (including Ollama-ready startup path)
  • Add CI workflow with branch/PR checks and tag-based UAT/PROD deployment jobs
  • Expand README with setup, usage, and deployment guidance

Change Size

  • 38 files changed
  • 2107 insertions, 2 deletions

Commit History Included

  • initial commit and subsequent foundational commits on API, RAG, ingestion, evaluation
  • CI/CD pipeline and deployment tagging
  • Docker/runtime stabilization and dependency updates
  • README/documentation expansions
  • latest config updates (Ollama timeout + Chroma telemetry disablement)

Validation

  • Docker build completed successfully for enterprise-chatbot.

@FairozaAmira
initial commit
77c821e
@FairozaAmira
Add response metadata and docs for interactive API testing
b4e951f 
- include timeTaken and apiCost fields across chatbot, crawler, and evaluation responses
- align cost estimation with optional LLM usage and update tests to assert new schema
- document Swagger/Redoc endpoints for local exploration in README
@FairozaAmira
Expand RAG documentation and sample datasets
57e151d 
- add sample knowledge base assets across knowledge, tender, finance, and mock web pages for crawler demos
- enhance data crawler to ingest local HTML “web samples” and update README with directory structure details
- document full request/response examples for crawler, chatbot, and evaluation APIs (including id/createdAt/timeTaken/apiCost metadata)
- alias response fields to camelCase in shared schemas to keep JSON output consistent
@FairozaAmira
Add external website ingestion and expand API documentation
c57afbc 
- allow crawler requests to supply website_urls and fetch external pages into the knowledge base
- update sample data, tests, and README with new crawler payload/response examples and recommended URLs
- document json response metadata and camelCase aliases for all endpoints
@FairozaAmira
Add CI/CD workflow with tag-based UAT and PROD deployments
36481f2 
- rename workflow to CI-CD and trigger on all pushes/PRs
- add deploy-uat and deploy-prod jobs that run on tag suffixes (*-uat, *-prod)
- each deploy job builds Docker image and includes placeholder deployment steps
- jobs reference GitHub environments (uat, production) for protection/secrets
@FairozaAmira
Fix setuptools flat-layout build error
30fac8f 
- configure explicit package discovery to include only the Python package
- add [tool.setuptools.packages.find] with include pattern
- resolves CI build failure caused by multiple top-level packages (data, routers, controllers)
@FairozaAmira
Fix LangChain import error and modernize RAG pipeline
297b770 
- replace deprecated RetrievalQA with load_qa_chain and manual retriever invocation
- update imports to use langchain_core and stable langchain.chains modules
- adjust RAGPipeline._build_chain and answer methods to new API
- resolves CI test failure: ModuleNotFoundError: No module named 'langchain.chains'
@FairozaAmira
Remove deprecated langchain.chains imports and simplify RAG pipeline
8ac94cd 
- drop all imports from langchain.chains to fix ModuleNotFoundError
- implement RAG using only langchain_core and direct LLM invocation
- format context from retrieved docs and invoke LLM with prompt
- ensures compatibility across recent LangChain versions
@FairozaAmira
Freeze Python and dependency versions; enforce CPU-only builds
e7d51fe 
- pin Python to 3.10.11 in pyproject.toml and Dockerfile
- pin all runtime and dev dependency versions
- add pydantic-settings and update BaseSettings import
- set CPU-only environment variables in Dockerfile to prevent CUDA/NVIDIA packages
- ensures reproducible builds and avoids GPU dependencies
@FairozaAmira
Fix BaseRetriever import and enforce CPU-only PyTorch in Docker
b919685 
- add missing BaseRetriever import from langchain_core.retrievers
- preinstall CPU-only torch before uv pip install to prevent CUDA/NVIDIA packages
- ensures tests import correctly and Docker builds stay CPU-only
@FairozaAmira
Fix Pydantic v2 private attribute error in DomainAwareRetriever
ea39d2e 
- use object.__setattr__ to set _base_retriever and _domain
- bypasses Pydantic field validation for private attributes
- resolves ValueError: object has no field "_base_retriever"
@FairozaAmira
Remove torch/torchvision from dependencies; enforce CPU-only in Docker
90ab967 
- remove torch and torchvision from pyproject.toml to avoid conflicts
- keep CPU-only torch installation in Dockerfile via pip3
- prevents dependency resolver from pulling CUDA-enabled torch
- ensures reproducible CPU-only Docker builds
@FairozaAmira
Add env-specific deploy configs and Ollama-ready Docker runtime
e98fff7 
- add .env.uat and .env.production templates to project root
- update .gitignore to keep .env ignored but track .env.uat/.env.production
- update CI deploy jobs to load .env.uat/.env.production before build
- add Ollama runtime setup in Dockerfile (including zstd requirement)
- add startup script to auto-start Ollama and pull configured model
- document env template usage and CI secret guidance in README
@FairozaAmira
Add env-specific deployment configs and stabilize Docker/Ollama runtime
20c6a64 
- add .env.uat and .env.production templates (tracked in repo)
- update .gitignore to keep .env ignored but allow env deploy templates
- update CI deploy jobs to load .env.uat/.env.production before Docker build
- add Ollama runtime bootstrap script and Docker support for model pull
- install zstd for Ollama installer in runtime image
- fix Chroma sqlite compatibility via pysqlite3 fallback + dependency update
- add README guidance for env templates, Ollama mode, and CI-safe usage
@FairozaAmira
Stabilize Docker/Ollama setup and deployment env templates
b9bd3e6 
- add tracked .env.uat and .env.production templates for CI deploy jobs
- update CI to load env-specific files before UAT/PROD Docker builds
- harden Docker runtime for Ollama (zstd install, startup script, model pull flow)
- fix Chroma sqlite compatibility in containers (pysqlite3 shim + builder deps)
- clean README Docker commands and env examples (no inline value comments)
@FairozaAmira FairozaAmira self-assigned this Feb 21, 2026
@FairozaAmira
update readme
3e0eab6
@FairozaAmira
build: switch to bookworm images and remove sqlite compatibility shim
21910e7 
- use python:3.10-slim-bookworm for both builder and runtime images
- remove apt install step for build-essential/libsqlite3-dev
- relax Python requirement from ==3.10.11 to >=3.10,<3.11
- remove Linux-only pysqlite3 dependency
- delete sqlite3 monkeypatch fallback in vectorstore
@FairozaAmira
feat(config): add configurable Ollama timeout and disable Chroma tele…
56e27b7 
…metry

- add `ollama_timeout_seconds` setting (default 300s) and pass it to `OllamaLLM`
- increase Ollama default timeout from 60s to 300s
- disable Chroma telemetry at app startup (`ANONYMIZED_TELEMETRY=False` + no-op posthog capture)
- update [.env.uat](cci:7://file:///Users/fairoza/Documents/Personal/sample-chatbot/.env.uat:0:0-0:0), `.env.production`, and README with new timeout/telemetry env vars
@FairozaAmira
FairozaAmira merged commit 8bdc299 into main Feb 21, 2026
6 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

@FairozaAmira FairozaAmira

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@FairozaAmira