2025-09-13 –, Track 2
FastAPI has taken the Python community by storm with promises of blistering performance but is it truly the fastest choice for your APIs? In this session, we’ll pull back the curtain on FastAPI’s async foundation, Pydantic powered validation, and auto‑generated docs to understand the real sources of its speed. Then, through benchmarks comparing FastAPI with Flask, Django, Express.js, and Go’s Gin you’ll see exactly where it excels, where it doesn’t, and why context matters more than hype. Walk away knowing when FastAPI is your secret weapon and when another framework might better serve your real‑world needs.
- Introduction & Motivation (5 min)
- Quick audience poll: FastAPI experience
- The “FastAPI hype” vs. real‑world questions
- Why performance matters in frameworks
- FastAPI Architecture Deep‑Dive (6 min)
- Starlette + uvloop + Uvicorn async core
- Pydantic validation impact
- Automatic OpenAPI/Swagger docs
- How each layer impacts request flow
- Benchmark Methodology (4 min)
- GitHub repo walkthrough: reproducible tests
- API Test categories
- Single Query
- JSON serialization
- Multiple Queries
- Data Update
- Plain Text
- Comparison targets: Flask, Django, Express.js, Gin
- Benchmark Results & Analysis (8 min)
- Latency and throughput charts
- Use‑case breakdown: I/O‑bound vs. CPU‑bound endpoints
- Key takeaways: where FastAPI leads and where it trails
- Framework Decision Guide (4 min)
- Balancing raw speed with ecosystem maturity, libraries, team expertise
- Ideal scenarios for FastAPI (microservices, real‑time apps)
- When to favor Flask/Django or alternative stacks
- Benchmarking Repo Walkthrough (3 min)
- Cloning and running the benchmark suite
- Customizing tests for your own workloads
- Interpreting results to inform your project decisions
- Q&A & Next Steps (5 min)
- Share GitHub repo and docs
- Further reading and community resources
- Open floor for audience questions
This is the link to my previous talk at Escape 2024 by BBD software : https://youtu.be/ohNsulYlgQE?si=G6iGmmNw0Z9LpU32
Please note PyCon2025 will be having new benchmarkings and totally new format of talk of previous talk.
python language
Additional Resources – Target Audience –Intermediate
Software engineer at BBD Software. Working closely with Django, FastAPI, Azure, GenAI real life use case implementations, AI Agents.
Curious, life long learner.