Hi, I'm
Saeed Khan
Senior Data Engineer · Dhaka, Bangladesh
I build data pipelines and platforms that hold up under real production load — then use AI tooling to make them safer to change.
About
Senior Data Engineer (6 yrs) who pairs deep data engineering skills with AI-agent automation — deploying and hardening agents for automating repetitive data-warehouse operations. Delivered concrete platform wins: a 31x query-latency reduction via Postgres indexing, and a ~15x storage-cost / ~5.8x latency gain migrating a 100GB+ raw-data layer to an object-storage lakehouse. Currently building a data-intelligence platform serving 300+ institutional clients.
Data platform architecture
Pipeline reliability & observability
Cloud cost & performance optimization
AI-assisted automation
Experience
Cefalo Bangladesh Ltd.
Client: a B2B data intelligence platform serving 300+ institutional clients
Dhaka, Bangladesh (Remote)
Senior Software Engineer I (Data Engineer) — Feb 2025 – Present
Software Engineer (Data Engineer) — Jan 2023 – Feb 2025
- Architected and scaled the platform's core data-ingestion layer across dozens of external registries and APIs, powering analyst dashboards, forecasting, and decision-support tooling.
- Cut latency and processing cost by partitioning large datasets for parallel ETL execution, reducing memory footprint on high-volume executor jobs.
- Designed a two-metric freshness monitoring system (write-lag vs. coverage-lag) with an interactive dashboard, closing a gap where dead sources still read 85% healthy — piloted on 10 pipelines with Slack alerting for proactive action on stale or late data.
- Eliminated full-table scans on the platform's largest table (172M rows) via 4 zero-downtime indexes — cut query latency up to 31x, from 832s to sub-second.
- Migrated a 100GB+ dataset from Postgres to an S3 Parquet + DuckDB lakehouse — ~15x cheaper storage, ~5.8x faster queries, validated to 85–100% byte parity.
- Iteratively hardened a production auto-fix agent's prompts and safety harness (~50 automated repairs), closing self-verification gaps to push unattended fix-rate toward 85% without manual intervention.
- Orchestrated all ETL workflows with Prefect; containerized jobs with Docker and built CI/CD pipelines with GitHub Actions for automated deployment to GCP Cloud Run.
- Built an internal AI HR agent (LangGraph, FastAPI, React, Ollama) with a RAG pipeline over policy documents and employee database integration.
- Mentored junior engineers on unit testing and pipeline design, and designed onboarding exercises for new team members.
Robi Axiata Ltd.
Bangladesh's second-largest telecom operator, 50M+ subscribers
Dhaka, Bangladesh
Specialist, Big Data & Analytics Development — June 2020 – December 2022
- Developed an automated data ingestion framework migrating terabytes of daily transactional data from Oracle to Hadoop via Apache Sqoop, integrating 20+ sources for batch and real-time processing.
- Built and maintained 30+ data marts covering 50M+ customers to serve Tableau analytics and BI reporting for business stakeholders.
- Enhanced ETL data quality pipelines for monitoring, validation, and job observability, reducing manual oversight.
- Implemented reusable Python ETL modules standardizing ingestion and transformation patterns across the analytics platform.
- Gained production experience with Hadoop, Hive, PySpark, Parquet/ORC, Apache Ranger, and Cloudera Data Platform.
Products & Systems
Features and systems I've built as part of the B2B data intelligence platform at Cefalo.
Projects
Personal and learning projects, built outside of work.
Skills
Certifications
Contact
Let's build something reliable together.
Open to data engineering and platform roles, and always happy to talk pipelines, architecture, or AI-assisted tooling.