Shaheer Bin Ijaz
Senior Software Engineer · Python · Django · FastAPI · ETL · Computer Vision · Blockchain
shaheer.ijaz03@gmail.com
Lahore, Pakistan
github.com/ShaheerIjaz97
linkedin.com/in/shaheer-bin-ijaz
Summary
Six years building Python services that take real production load. The work sits between three jobs companies usually hire for separately. A Django service talking to a payment provider one day. A Lambda pipeline ingesting product data into BigQuery the next. A YOLO model checking parts on a manufacturing line the day after. The skill is treating all three as the same job: a system that runs in prod, gets observed, gets fixed when it breaks.
Experience
-
Senior Software Engineer · Devsinc
Aug 2024 — Present
Lahore, Pakistan · On-site
- · Lead backend on Kenya's national healthcare initiative (HealthPro ERP / Tiberbu) and Electus Education Global, transitioning monolithic Django systems into FastAPI microservices for faster deployment cycles.
- · Built event-driven systems with Google Pub/Sub, RabbitMQ, and the outbox pattern to guarantee reliable async communication across education and healthcare services.
- · Run GCP infra (Cloud Run, VPCs, Scheduler) plus AWS on Tiberbu, with CI/CD pipelines built for compliance on mission-critical national healthcare workloads.
- · Tune PostgreSQL via connection pooling, query plans, partitioning, horizontal scaling to hold latency under increased education and healthcare load.
-
Python Software Engineer · Code District
Jun 2022 — Aug 2024
Lahore, Pakistan
- · Built the Refund Management System from scratch in Django REST Framework. State-machine workflow, immutable audit trail, role-based approvals. Public reconstruction is on my GitHub.
- · Designed and shipped a Solana NFT marketplace backend in FastAPI, with off-chain Postgres read model and an indexer that follows on-chain events.
- · Integrated Solana on-chain reads, payment flows, third-party auth across the marketplace stack.
- · Built repeatable internal patterns: schedulers, seeders, custom user models, custom commands, middleware.
-
Python ETL Developer · Techesthete
May 2021 — Jun 2022
Lahore, Pakistan · On-site
- · Built ETL pipelines pulling product data from Walmart and Instacart APIs, transforming through AWS Lambda, loading into Google BigQuery. Public reconstruction on my GitHub.
- · Developed Django REST Framework backend APIs sitting on top of the warehouse.
- · Owned data-quality checks and pipeline observability for high-volume retail feeds.
-
Machine Learning Engineer · WiMetrix · Industrial Data Intelligence
Nov 2019 — Apr 2021
Lahore, Pakistan
- · Wrote and tuned ETL queries against SQL Server, extracting and merging operational tables for downstream ML.
- · Trained CNN-based YOLO models to catch fabric faults on a live textile manufacturing line. Public reconstruction on my GitHub.
- · Built regression models predicting worker efficiency from operational signals.
- · Solved time-series forecasting with LSTMs to predict shift-level output efficiency.
Stack
- Languages:
- Python, SQL, TypeScript (where needed)
- Backend:
- Django, DRF, FastAPI, Celery, SQLAlchemy 2, Pydantic v2
- Data:
- Postgres, BigQuery, Pandas, NumPy, DuckDB
- ML / CV:
- PyTorch, YOLOv8, OpenCV, Keras, scikit-learn
- Cloud & infra:
- GCP Cloud Run, AWS Lambda / S3, Docker, Nginx, Redis
- Event-driven:
- Google Pub/Sub, RabbitMQ, outbox pattern, Celery
- Blockchain:
- Solana RPC, solders, off-chain indexers
- Observability:
- OpenTelemetry, Prometheus, structlog
Education
-
Master's degree, Computer Science · University of Engineering and Technology, Lahore
2020 — 2022
-
Bachelor of Science, Computer Sciences · Lahore Garrison University
2015 — 2019
Engineering principles
- · Schema first. Boundaries get explicit types.
- · State machines over CRUD. Anything with a lifecycle (refunds, orders, indexer events, training runs) gets a state machine with explicit allowed transitions.
- · Idempotent writes. Anything that hits a downstream sink (warehouse, blockchain indexer, payment provider) uses a deterministic key plus an upsert.
- · Observability is not optional. Every service ships with structured logs, healthz, readyz, plus metrics that say something useful (rows in, rows rejected, latency p95).
- · The architecture goes in a doc. Every repo of mine has an ARCHITECTURE.md explaining the load-bearing decisions plus the tradeoffs that came with them.