Week #4

Week #4 #

Testing and QA #

Strategy & types of tests implemented

  • Manual endpoint testing via Swagger UI and Postman:
    • Verified /api/poi search and /api/poi/recommendations endpoints.
    • Checked authentication (/api/token/get-token, /api/token/refresh, /api/token/current-user).
  • Data‐quality validation on a hand‐labeled set of 60 real-world queries.
  • Performance measurements with three custom scripts:
    • measure_index_load.py (index load time & memory),
    • measure_search_latency.py (SBERT.encode vs FAISS.search latency),
    • compute_precision.py (precision@5/10).

Evidence of test execution #

  • Index load: 6 ms, RSS ↑ from 103.84 MB to 106.16 MB
  • Latency (60 queries):
    • encode_time_ms: mean ≈ 92.6 ms
    • search_time_ms: mean ≈ 0.77 ms
  • Accuracy on validation set (top-5 judgments):
    • precision@5 ≈ 0.84
    • precision@10 ≈ 0.76

All raw results are in DLS/experiments/results/*.csv.

CI/CD #

  • Current status: No CI/CD pipeline yet.
  • Tools planned: GitHub Actions + Docker Compose.
  • Next steps:
    1. Add a GitHub Actions workflow to build the backend image, run lint/tests, and push to a container registry.
    2. Integrate frontend build and unit tests into the pipeline.

Deployment #

Staging #

  • Local staging via docker-compose up --build (services: backend, Postgres).
  • Configuration variables in .exampleenv; mounts include code and DLS data.

Production #

  • Not yet implemented — planned for next sprint.
  • Goals:
    • Cloud VM or Kubernetes deployment.
    • Domain name, SSL, environment configs, autoscaling.

Vibe Check #

  • Progress: Core search & recommendation service working; experiments completed.
  • Roadblocks: CI/CD missing;
  • Team dynamics: Emil led backend/DLS; Vlad drove frontend integration;

Weekly commitments #

ParticipantContributionLink to Commit
Emil• Implemented FAISS‐based search service, query encoding, alcohol filtering
• Wrote scripts for latency & precision measurement
• Authored documentation & Docker setup		• https://github.com/IU-Capstone-Project-2025/EasyTravel/pull/19
• https://github.com/IU-Capstone-Project-2025/EasyTravel/pull/20
• https://github.com/IU-Capstone-Project-2025/EasyTravel/pull/21
Vlad• Built frontend UI for search and recommendations
• Integrated with backend API & auth
• Resolved UI bugs and styling issues		• https://github.com/IU-Capstone-Project-2025/EasyTravel/commit/dd1efbaae6f44ca0a654affbf98670556c934a6c
• https://github.com/IU-Capstone-Project-2025/EasyTravel/commit/856bf37d8f61022ac76043867366b9b832463214
• https://github.com/IU-Capstone-Project-2025/EasyTravel/commit/a04cd3ca9e2ed0c82cbaedce3af0f86ec8f621ef

Plan for Next Week #

  1. Write unit & integration tests for backend.
  2. Configure CI pipeline (GitHub Actions).
  3. Finalize production deployment docs & scripts.
  4. Harden frontend, add E2E tests.

Confirmation of the code’s operability #

We confirm that the code in main branch:

  • is in working condition
  • can be started via docker-compose up --build (see Backend/README.md)