AI agents that do the work — with the data-quality automation that keeps them honest.
I build AI agents — starting with a production text-to-SQL analyst agent, and more on the way — plus the pipelines, data-quality checks and dashboards that make the numbers trustworthy in the first place.
Services
What I do
Dashboards & reporting
Power BI and Looker Studio. I write the metric definitions down first, so a number means the same thing in every report.
Data-quality checks
Automated checks that catch records which look fine but aren't — with the impact broken down by cause, not just a vague warning.
Pipelines & scheduling
ETL on Airflow or Cloud Composer into BigQuery. Once it runs, the reporting layer updates itself and nobody has to remember to.
AI analyst agents
An agent that writes its own SQL, checks it before running it, then explains the result on your own machine — no data row leaves the building.
Skills
What I work with
Works
Selected work
AI Data-Analyst Agent (Text-to-SQL)
An agent that runs the whole analysis on its own and answers free-form questions — writing its own validated SQL, with results explained locally so no data row leaves the machine.
Automated Data-Quality Framework
A 15-check framework that finds mislabelled "successful" VPN sessions and explains a success-rate drop check by check, instead of leaving it a mystery.
Mobile App Analytics Dashboard
Acquisition, retention, revenue and quality in one interactive dashboard, fed by a modelled BigQuery layer rather than raw event tables.
Automated ETL Pipeline on GCP
A five-DAG scheduled pipeline turning a raw CSV into an analysis-ready BigQuery table — idempotent loads, per-stage retries, no manual steps.
Customer Behaviour Analytics
Ten commercial questions answered in SQL and a Power BI dashboard that showed the real opportunity was monetising an already-loyal base.
Resume
Education & experience
Education
Distinction
Distinction
First Class Honours
Experience
VPN product telemetry across six apps — Looker Studio dashboards, a 15-check data-quality framework and an AI analyst agent.
Power BI reporting across five departments, an asset-risk data audit, and cross-team operational logging.
Financial-data reconciliation and regulatory reporting on Azure/Databricks with SQL and Python for audit clients.
Built an automated ETL pipeline on Cloud Composer / Airflow loading 16,600 records into BigQuery.
Recommendations
What colleagues say
“Asghar is a true multitasker — but he makes sure every single task is done phenomenally. The effort he puts in is evident in his work; he never hands anything over half-done.”
“Asghar turns a vague data problem into a working solution fast. He thinks like an engineer and an analyst at the same time — the text-to-SQL agent he built is genuinely production-grade, not a proof-of-concept.”
“I've watched Asghar take on problems well outside his comfort zone and still ship something clean and well-reasoned. He's meticulous with the details and dependable when it counts — exactly who you want owning your data.”
Pricing
How to work with me
Indicative rates, benchmarked to Upwork's 2026 averages for data & AI work. Final price depends on scope.
Hourly · Ad-hoc
- One-off analysis or a fix
- A dashboard or agent review
- Pay for the hours used
Fixed project
- A dashboard, pipeline or AI agent, end to end
- Deliverables and price agreed up front
- Metric definitions written down
Retainer · Monthly
- Ongoing reporting, data quality & agents
- A set number of days each month
- Priority on new requests
AI Insights
Latest in AI agents
From “what happened” to “what's next”
Dashboards that only report are giving way to agents that investigate why a metric moved, flag what's coming, and route the right alert to the right person — continuously, with no SQL query in sight. It's the exact shift the HEXA AI analyst agent I built is designed around.
Read more →Data analysis is the #2 agent use case
Over 57% of enterprises now run AI agents in production, and research & data analysis is the second most common use case (24.4%) in LangChain's 2026 State of AI Agents survey — the same LangGraph stack behind my own text-to-SQL agent.
Read more →Agents that query data where it lives
New agentic database features — like Oracle's 2026 AI Database — let agents securely reach real-time enterprise data in place and reason over it with an LLM. It's the privacy-first pattern my agent uses so no data row ever leaves the machine.
Read more →Contact
Let's talk
I'll ask a couple of questions, then come back with a first draft.