Ravi Chandran Senior Python Developer
Seattle, WA • pydev@gmail.com • +1 (206) 555 0184
Profile Summary
- Senior Python Developer with 8 years of experience building Python back-end services at consumer scale across visual discovery, ads ranking, and creator platforms, specializing in FastAPI services, asyncio concurrency, and SQLAlchemy 2.x data access.
- Hands-on coverage across Python (Python 3.12), web (FastAPI), ORM (SQLAlchemy 2.x, PostgreSQL), async (asyncio), and containers (Kubernetes, AWS) with strong fundamentals in PEP 8 discipline, full type hints, dataclasses, comprehensions, generators, and context-manager idioms.
- Deep expertise in async/await and asyncio orchestration, REST and GraphQL API design, Pydantic validation and dependency injection, and cProfile and py-spy performance tuning, applying methodologies such as test-driven development with pytest fixtures and parametrize and strict typing with mypy and ruff on every PR to deliver reliable, well-tested Python services that hold up at consumer-app scale.
- Engaged collaborator working cross-functionally with Product, Data Science, ML, and Platform teams in trunk-based and review-heavy product teams, contributing to API design forums, on-call rotations, and architecture reviews with an ownership-first mindset and clean handoffs.
- Mentor who shares technical excellence and fosters a culture of type-safe, test-covered Python and PEP-driven code quality through PR reviews and pattern docs, while running Python guild and API design review sessions and authoring widely used service and data-access templates.
Technical Skills
- Languages & Runtime:
- Python 3.12, Python 3.11, Python 3.10, type hints (PEP 484/604), dataclasses, generators, comprehensions, context managers, the GIL, asyncio
- Web Frameworks & APIs:
- FastAPI, Django, Flask, Litestar, Starlette, Pydantic, REST, GraphQL with Strawberry, OpenAPI, middleware, dependency injection
- Async & Concurrency:
- asyncio, async/await, anyio, trio, async generators, multiprocessing, threading, concurrent.futures, Celery, Dramatiq, httpx, aiohttp
- Data Access & ORM:
- SQLAlchemy 2.x, Django ORM, Tortoise ORM, SQLModel, Alembic, PostgreSQL, MySQL, SQLite, Redis, MongoDB, DynamoDB
- Testing & Quality:
- pytest, pytest-asyncio, fixtures, parametrize, unittest, hypothesis, responses, respx, Testcontainers, factory_boy, freezegun, coverage.py
- Tooling & Packaging:
- uv, Poetry, Hatch, pip-tools, ruff, Black, isort, mypy, pyright, pre-commit, tox, nox
- Data, ML & Scripting:
- pandas, polars, NumPy, PyArrow, scikit-learn, PyTorch, Hugging Face, Jupyter, ETL pipelines, Airflow, Prefect
- Deployment & Cloud-Native:
- Docker, multi-stage builds, distroless, Kubernetes, AWS Lambda, GCP Cloud Functions, Gunicorn, Uvicorn, Hypercorn, GitHub Actions, AWS
Education
University of
Washington
B.S. in Computer
Science
Work Experience
Pinterest
Senior Python
Developer
- Owned idiomatic Python service development end to end on the creator and ads back-end platform serving 480M monthly users, shipping creator API, ads ingestion, and moderation pipeline across 14 type-hinted Python services held to PEP 8 and the Zen of Python.
- Tuned async and concurrency with asyncio task groups with bounded semaphores and httpx clients, structured cancellation on every request path, and multiprocessing for CPU-bound jobs that the GIL would otherwise pin, capping concurrent tasks at 1024 and pulling fan-out latency from 2.3s down to 340ms on the creator-feed aggregation path.
- Built backend and web services with FastAPI with Pydantic models and dependency-injected services, typed request and response schemas, middleware for auth and rate limiting, and OpenAPI specs generated from code across 180+ endpoints, dropping p95 from 220ms to 48ms on the hot creator-lookup path.
- Designed APIs and third-party integrations with REST with httpx clients, webhook fan-out, and OpenAPI contract testing, retries with exponential backoff, idempotency keys, and respx-mocked CI checks across 22 partner integrations, sustaining webhook throughput of 40k events/min with zero duplicate-delivery incidents over the last four quarters.
- Owned data access and ORM work with SQLAlchemy 2.x async sessions with Alembic migrations and Redis caching, typed query builders, eager-loading audits, and prepared statements on hot reads, shipping 160+ schema migrations and cutting query p99 on the ads-attribution path by 71%.
- Drove performance and profiling work with cProfile and py-spy on hot paths with scalene memory profiling, lru_cache and Redis layered caching, vectorized NumPy in inner loops, and Cython for the two hottest scoring kernels, dropping memory footprint per worker from 1.4GB to 310MB and cutting CPU cost on the ranking pipeline by 58%.
- Shipped deployment and cloud-native Python with multi-stage slim Docker images with Uvicorn workers on Kubernetes, AWS Lambda for spiky webhook fan-out, Gunicorn-managed worker pools, and GitHub Actions pipelines on Kubernetes releasing multiple times daily, shrinking container images from 920MB down to 148MB across the fleet.
Zillow
Python
Developer
- Drove testing and code quality with pytest with fixtures, parametrize, and Testcontainers integration, hypothesis property tests on parser code, respx-mocked HTTP boundaries, and coverage gates in CI, lifting test coverage from 44% to 89% and cutting production bugs per release by 64%.
- Modernized Python tooling with Poetry lockfiles, ruff, Black, and mypy on pre-commit, tox matrices for 3.10 and 3.11, and pip-tools resolution gates on dependency PRs, dropping CI pipeline runtime from 18min to 4min and lifting mypy type coverage across the monorepo to 92%.
- Built data and ML scripting workflows with pandas and polars ETL jobs feeding scikit-learn pricing models, PyArrow for columnar interchange, Airflow DAGs for nightly batches, and Jupyter notebooks converted to production scripts, cutting ETL runtime by 73% while processing 180M rows nightly across the home-valuation pipeline.
- Strengthened idiomatic Python foundations across the team through dataclasses, generators, and context-manager refactors, package-layout reviews, and weekly PEP reading sessions, while mentoring 7 engineers through pair-programming, code-review checklists, and a written Python idiom guide adopted across two squads.