ProjectIntent Over Keywords: From Stitched Embeddings to a Production Fashion Agent · Module 2: The stitched architecture · part 4 of 26
Part 4 · Module 2: The stitched architecture

Module 2: The stitched architecture

23 min read·
About 50 minutes · Needs: Module 1 (the problem with multimodal search, the test video) · Code: part1-video-rag-stitched/src/api/main.py, src/ingestion/pipeline.py, src/retrieval/retriever.py, src/api/uploads.py, originals/v1/video-rag-old/src/

Where this fits

WHERE THIS FITS Intent Over Keywords this module PART 1: STITCHED VIDEO RAG three models, two spaces, RRF PART 2: UNIFIED VIDEO RAG one model, one space PART 3: STYLEME FastAPI, LangGraph, React QUALITY AND OPERATIONS evals, Docker, Cloud Run Frames ffmpeg, one per 5 s CLIP ViT-B/32 frames + text, 512 d Whisper base speech to text MiniLM 64-word chunks, 384 d Two collections p1_visual, p1_text Rank fusion RRF, k = 60 Gemini answer from transcript text API + Streamlit upload, ask, sources Windows ffmpeg, video + audio Embedding 2 Gemini, one vector One collection p2_unified, no fusion Notes + answer Gemini, cited times React web chat, photo, chips nginx one URL, /api proxy FastAPI /api/chat, limits Sessions in memory, cap of 3 Planner intent before retrieval Implementer batched embed, search Generator stylist reply, ids Grounding check no LLM, drops invented Gemini flash-lite, embedding 2 Qdrant styleme_products Catalogue ingest image + title vectors Langfuse optional traces Test video 26 questions, 4 kinds Video eval Part 1 vs Part 2 Agent evals intent, grounding Guards uploads, limits, secrets Containers compose profiles Cloud Run StyleMe only CI/CD GitHub Actions, WIF Original code v1, v2, fashion-agent
This module (highlighted): all of Part 1. Blue is our system, amber a model (CLIP, Whisper, MiniLM, the Gemini answer model).

This module is the whole of Part 1 at once: how a video goes in, how a question comes out as an answer, and which code connects the pieces. Modules 3 to 6 then open one box each (frames and CLIP, speech, the two collections and rank fusion, the failure). Here you learn the shape of the system, the contract it shares with Part 2, and how the author's original code was hardened without changing its design.

The rest of this project is yours to keep

This project is bought on its own, once, and stays readable afterwards, including the parts added to it later.

Buy this project · ₹799See what is inside

Or open everything with a plan