AWS & GCP certified data engineer

Mesum Asghar

I’m a data engineer with a passion for agentic AI and machine learning. With two years of experience building pipelines, warehouse models, and automated workflows, I turn raw data into reliable systems and explore how intelligent agents can make them more useful.

Hire me on Upwork View selected work
  • AWS
  • GCP
  • Snowflake
  • LangGraph
  • Spark
  • Data Engineering · dbt, Airflow, Spark
  • Warehouses · BigQuery, Snowflake, Redshift
  • Machine Learning · scikit-learn, XGBoost
  • Agentic AI · LangGraph, tool calling, RAG
Portrait of Mesum Asghar

Selected work

Systems with a clear job.

Research agents, retrieval apps, and warehouse pipelines. Each one is built to be inspected, not just demoed.

01 Applied AI

Verity

A self-correcting research agent. It drafts an answer, critiques what is missing, searches the live web, and revises the result with citations.

  • LangGraph
  • OpenAI
  • Python
View on GitHub
02 Applied AI

Totoro

A conversational companion grounded in a curated mental-health knowledge base. Follow-up questions are rewritten for search, and crisis language is routed to lifeline resources.

  • LangChain
  • Pinecone
  • Streamlit
View on GitHub
03 Analytics

Retail Payment Intelligence

A retail analytics pipeline on synthetic e-commerce data. Medallion layers take raw payments through to Looker Studio dashboards.

  • dlt
  • dbt
  • BigQuery
  • Airflow
View on GitHub
04 Batch

AWS Spark Batch

A batch pipeline that lands S&P 500 ETF data in S3, transforms it with Spark through medallion layers, and serves it in Superset.

  • Spark
  • Airflow
  • S3
  • Superset
View on GitHub
All projects

Services

From raw sources to a number someone can defend.

01

Pipelines

Ingestion with validation, incremental loads, backfills, CDC, and reverse ETL.

02

Models

Star schemas and medallion transforms, with tests, lineage, and a Gold layer built for questions.

03

Warehouses

BigQuery, Snowflake, and Redshift layouts, plus dashboards a non-technical stakeholder can open.

04

APIs & AI

FastAPI services, webhooks, and LangChain or LangGraph agents wired into the data they need.

Tools

The stack I ship with.

Languages

Python, SQL, JavaScript, TypeScript, Java, C

Data Engineering

dbt, dlt, Airflow, Kafka, Spark, change data capture

Databases & Warehouses

BigQuery, Snowflake, Redshift, PostgreSQL, data modeling, query optimization

Machine Learning

pandas, NumPy, scikit-learn, XGBoost, feature engineering, model evaluation

LLM Applications

OpenAI API, embeddings, RAG, Pinecone, retrieval with citations

Agentic AI

LangChain, LangGraph, tool calling, stateful workflows, human-in-the-loop approval

APIs & Cloud

FastAPI, Docker, GCP, AWS S3

Analytics & Dashboards

Looker Studio, Apache Superset, Streamlit

Contact

Tell me what the data needs to do.

A pipeline, a model, a dashboard, or an agent on top of data you already have. Send the context and I will reply with a way to start.

Email LinkedIn GitHub Scholar Upwork