Скачать или смотреть 2025 06 10 Project 1 Session 1 - TDS May 2025

Covered project overview and 5 key steps: data sourcing (Anand’s repo, Discourse API), preparation (embeddings), app creation, deployment, and testing.
Explained scraping (BBC weather app, network tab, JSON, auth via cookies, undetected ChromeDriver).
Discussed GitHub setup, FastAPI, curl, YAML config, testing, prompt engineering, LLM rubric.
Deep dive: Discourse API calls, cookie-based auth, and range-based data filtering.
Note: PDSA session tomorrow.


00:00:01 Introduction
00:04:43 Sourcing the Data
00:05:17 Clarification on Questions Outside Range
00:08:32 Preparing the Data
00:11:37 Creating the Application
00:13:30 Deployment
00:14:20 Testing the Deployment
00:16:05 Questions and Answers
00:24:20 Sourcing Data
00:24:50 Public Repos and Course Content
00:26:13 Discourse Scraping using API
00:35:02 API call Demonstration
00:39:13 Code demonstration
00:48:36 Authentication and API
00:50:52 Break and Questions
00:55:25 Polls and Project Discussion
00:57:03 Environment Variables for API Keys
01:00:23 API Key and Application Structure
01:05:34 Project Details Discussion
01:09:15 Discussion on different models
01:19:22 Authentication and API
01:22:11 Discourse Data Scraping
01:27:57 Cookie Usage and Application Flow
01:34:30 Questions and Answers
01:39:07 Vercel and Deployment Issues
01:43:14 Promptfu Testing
01:56:31 Questions and Answers
02:00:43 Cookie usage
02:04:19 API and Topic Identification
02:19:04 Data Processing for Embedding
02:25:56 Discussion on number of chunks to return
02:26:33 Data Processing and Embedding
02:28:05 Thank you and Closing