Avee Chakraborty’s Post

FastAPI Backend Journey 🚀 | Day 1 I started learning FastAPI thinking: “Okay… just another framework 😅” But when I actually used it… Everything felt fast, clean, and less confusing. Instead of writing long backend code, I was able to create my first API in just a few minutes. But I also made mistakes 👇 ❌ Messed up project structure ❌ Didn’t understand async properly ❌ Random file placements Then I realized something important: 👉 FastAPI is easy… but using it properly is what matters. So I decided to learn it step by step — not just coding, but how to structure it like a real backend developer. Starting today, I’ll share: 🔹 How to actually use FastAPI (not theory) 🔹 Proper folder structure 🔹 Real-world practices 🔹 Mistakes to avoid If you're learning backend like me, follow along 🤝 Tomorrow → Project Setup + Folder Structure #FastAPI #BackendDevelopment #Python #BuildInPublic #LearningJourney #Developers #100DaysOfCode #TechIndia

🚀 FastAPI: Why It’s Becoming the Go-To Framework for Modern APIs I used to think FastAPI was just another Python framework… But after working with it, I realized it’s built for how APIs should be designed today. Here’s what makes it stand out 👇 🔹 High Performance Built on ASGI with async support, FastAPI handles high concurrency efficiently — making it ideal for real-time and scalable systems. 🔹 Automatic Validation Using Python type hints + Pydantic, it validates requests and responses automatically. Less boilerplate, fewer bugs. 🔹 Auto-Generated Docs Swagger & ReDoc come out of the box. You get interactive API documentation without extra effort. 🔹 Async Support (When Needed) Perfect for I/O-heavy tasks like APIs, DB calls, and external integrations. But also flexible enough to work synchronously. 🔹 Developer Experience Clean syntax, strong typing, and easy debugging — makes development faster and more structured. 💡 What stood out to me: FastAPI isn’t just about speed — it’s about writing clean, reliable, and scalable APIs with less effort. In a world where performance and scalability matter, FastAPI is definitely worth exploring. Curious — have you tried FastAPI in production yet? 🤔 #FastAPI #Python #BackendDevelopment #APIs #Developers #Tech #LearningJourney

I used to think backend development was complicated. Too many concepts. Too much setup. It felt difficult to move from theory to actually building something real. Then I started learning FastAPI. This time, I didn’t just watch tutorials. I focused on building. - My Journey with FastAPI As a Python developer, FastAPI felt simple, fast, and practical from the start. It helped me understand how backend systems actually work instead of just learning concepts. - Why FastAPI? - High performance - Clean and easy syntax for Python developers - Built-in validation using Pydantic - Automatic API documentation with Swagger UI - What I’ve learned - Building REST APIs using GET and POST methods - Handling request and response data - Using Pydantic for data validation - Testing APIs using interactive Swagger documentation - What I’ve built I built backend projects using FastAPI where I created API endpoints, validated data, and tested everything through auto-generated docs. This hands-on experience helped me connect theory with real-world development. Working with FastAPI has improved my confidence in backend development and made me more comfortable building practical applications. If you are learning development and feel stuck, try building something instead of only consuming content. I would love to know what you are currently working on. #FastAPI #Python #BackendDevelopment #WebDevelopment #LearningInPublic

My CLAUDE.md was 281 lines. Claude loaded all of it every session - whether I was writing a post or debugging a Python script. Post rules. Format guides. Note conventions. Voice guidelines. All in context, all the time. Even when none of it was relevant. Longer context = lower adherence. The docs say it plainly. I ignored it until I ran this prompt: --- *Review my CLAUDE.md against https://lnkd.in/g6UNtsC7 and suggest specific improvements - what to add, what to move to .claude/rules/ files, and what to cut.* --- Claude read the official docs, compared them against my file, and came back with a diagnosis. The biggest one was `.claude/rules/` - a directory most people don't know exists. Here's what it told me to do: Rules in `.claude/rules/` load just like CLAUDE.md - but you can scope them to specific file paths. ``` paths: - "posts/**" ``` That one frontmatter block means post rules only enter context when Claude is working inside that folder. Format guides, note conventions - same thing. Zero tokens wasted when they don't apply. After the restructure: - CLAUDE.md: 281 lines → 69 lines - Always-loaded context: ~180 lines total - 3 rule files load on demand, only when relevant The path-scoped rules are in the official docs at https://lnkd.in/g6UNtsC7. Almost nobody uses them. Run `/memory` in Claude Code to see exactly what's loaded in your current session. If it's over 200 lines, that's where the drift is coming from. What rule do you keep repeating to Claude because it keeps forgetting? #ClaudeAI #AITools #BuildInPublic #ProductManagement #DeveloperTools

Stop writing manual validation logic In traditional frameworks, you spend a lot of time writing code like: if not data.get("email"): raise ValueError... With FastAPI, you stop writing "checks" and start defining Schemas. By using Pydantic models, FastAPI does the heavy lifting for you: ✅ Automatic Parsing: Converts incoming JSON directly into Python objects. ✅ Data Validation: If a user sends a string where an integer should be, FastAPI catches it instantly. ✅ Clear Errors: It sends a detailed 400 error back to the client automatically—your function logic doesn't even have to run. The result? Cleaner code, fewer bugs, and a backend that "just works." Check out the snippet below to see how 5 lines of code can replace dozens of if/else statements. #Python #FastAPI #Pydantic #WebDevelopment #Backend #CleanCode

🚀 Day 55 – Exploring FastAPI (Modern Backend Magic!) Today I dived into FastAPI, one of the fastest and most efficient web frameworks for building APIs with Python. ⚡ 💡 What is FastAPI? FastAPI is a modern web framework that helps you build APIs quickly using Python, with automatic validation, documentation, and high performance. 🔥 Why FastAPI stands out: ✔️ Super fast (built on ASGI & Starlette) ✔️ Automatic API docs with Swagger UI 📄 ✔️ Type hints = better code + fewer bugs ✔️ Easy to learn and implement ✔️ Async support for high performance 🛠️ What I learned today: 🔹 Creating a basic API 🔹 Handling GET & POST requests 🔹 Path & Query parameters 🔹 Request validation using Pydantic 🔹 Auto-generated interactive docs 💻 Simple Example: from fastapi import FastAPI app = FastAPI() @app.get("/") def read_root(): return {"message": "Hello World 🚀"} 📌 Key Takeaway: FastAPI makes backend development simple, fast, and production-ready with minimal code. Consistency is the real power 💪 #Day55 #FastAPI #Python #BackendDevelopment #APIs #100DaysOfCode #LearningJourney 🚀

One pattern that changed how I build FastAPI backends: Stop returning raw database models from your endpoints. When your API response mirrors your ORM model 1:1, you're creating tight coupling between your database schema and your API contract. One schema change can break every client. The fix: dedicated Pydantic response models per endpoint. Here's what you get: 1. Auto-generated OpenAPI docs that actually match your responses 2. A clear data boundary - internal fields stay internal 3. Freedom to refactor your DB without touching your API contract Bonus: Pydantic's model_validator and computed fields let you shape responses exactly how your frontend needs them - no extra serialization logic scattered across your codebase. What patterns have saved you the most headaches in your backend work? #Python #FastAPI #WebDevelopment #SoftwareEngineering #FullStackDeveloper

"I didn't do that!" On one of my projects, I've been using ty from Astral (the folks behind ruff and uv). But recent versions started throwing errors that others have also reported as false positives. So I decided to switch back to mypy. I launched and asked Claude to replace ty with mypy. It updated pyproject.toml and scripts/lint.sh. Ran "uv sync". Clean swap, no issues. Then it ran the lint and reported back: "14 errors across the whole codebase, all pre-existing. None in my new code." 😂 Even AI says, "It wasn't me, it was like that when I got here." Some things are universal. #AI #DeveloperExperience #Claude #Python #mypy #ty

I just built a fully working RAG (Retrieval-Augmented Generation) system from scratch. 🚀. Here's what it does and how: 𝗪𝗵𝗮𝘁 𝗶𝘁 𝗱𝗼𝗲𝘀: Upload any PDF → ask questions → get answers with cited sources 𝗛𝗼𝘄 𝗶𝘁 𝘄𝗼𝗿𝗸𝘀: 1. PDF is loaded and split into 500-character chunks 2. Each chunk is converted into embeddings (vectors) 3. Stored in ChromaDB — a vector database 4. At query time, the top 3 relevant chunks are retrieved 5. Chunks + question are sent to a local LLM 6. Answer returned with source citations 𝗡𝗼 𝗵𝗮𝗹𝗹𝘂𝗰𝗶𝗻𝗮𝘁𝗶𝗼𝗻: The LLM is explicitly instructed to answer only from the retrieved context. If the answer isn't in the document, it says so — no fabrication. 𝗧𝗲𝗰𝗵 𝘀𝘁𝗮𝗰𝗸: Python · LangChain · ChromaDB · Sentence Transformers · Ollama · Streamlit Runs fully locally — zero API costs. GitHub: https://lnkd.in/drfgqa-B #AIEngineering #RAG #Python #LLM #MachineLearning #BuildingInPublic

Episode 04 of my Claude Certified Architect course is live — and this one is all code. Three episodes of theory. Agentic loops. Multi-agent coordinators. Structured context passing. Claim-source mappings. Session management. All of it. One Python file. Copy, paste, run. Here's what the file demonstrates: → A proper agentic loop that exits on stop_reason == "end_turn" — not an iteration cap (Ep 01 pattern) → A coordinator spawning subagents via a custom delegation tool — the correct raw SDK pattern (Ep 02) → Structured findings with source_url, retrieved_at, and confidence passed between every agent handoff (Ep 03) → A report writer that cites every claim with its source — attribution preserved end-to-end One thing I want to be honest about: the real Task tool for spawning subagents is a Claude Code / Agent SDK feature — it doesn't exist in the raw Anthropic Python SDK. The correct raw-SDK pattern is exactly what this file shows: define your own delegation tool and handle it yourself. The exam tests the architectural concept, not the specific tool name. That distinction matters. Total cost per run: under $0.01. Uses claude-haiku-4-5. No external APIs needed — the knowledge base is stubbed out so you can run it immediately. This is the capstone for Module 1 (Domain 1, 27% of the exam). Link in comments 👇 #ClaudeAI #AIAgents #MultiAgentSystems #PythonAI #LLMEngineering #AnthropicClaude #AIArchitecture #AIEngineering #GenerativeAI #AIDevTools