Qamar Zia · Independent AI & data consultant
I build production AI for regulated industries, and audit it for investors.
Since 2020 I’ve built churn, forecasting and document-AI systems for retail banks, a US healthcare revenue-cycle group and a sovereign wealth fund’s research desk. Venture investors also bring me in to review AI startups before they invest.
- Since 2020
- Building machine learning that runs in production
- 4 sectors
- Banking, healthcare, asset management, venture capital
- 20+ certifications
- IBM and Microsoft, including Azure Data Scientist (DP-100)
- 3+ years
- With a sovereign wealth fund’s bonds and FX research desk
Selected work
Systems I’ve built and reviewed
Client names are withheld under confidentiality. Each project links to a short write-up.
Cross-asset research platform
Pipelines and Dash apps that give a sovereign wealth fund’s bonds and FX researchers one place to pull, compare and backtest market data.
In daily use by research and trading for 3+ years
Retention & cash analytics
Churn models, customer segments and branch cash forecasts for retail banks, from feature store to monitored production.
Led a team of four across several banks
Claim-denial prediction
A model that flags insurance claims likely to be denied before they are submitted, served to internal apps through a monitored API.
Ran in production with monitoring
Support assistant in Microsoft Teams
A Teams and telephony assistant that answers support questions from the company’s own documents, using an OCR and entity-extraction pipeline underneath.
Live in a corporate Teams tenant
AI due diligence for investors
Technical reviews of early-stage AI startups for venture investors, plus retrieval tools that turn a data room into a first-draft investment memo.
Seed and Series A deals
From LLM wrappers to agents
Design reviews and workshops for venture-backed engineering teams moving from a single LLM call to agents built on LangGraph, CrewAI or AutoGen.
Several portfolio companies
Services
What I build
Six kinds of system, each drawn from projects that went into production. Select one to see how it works.
Assistants that work inside tools your teams already use, such as Microsoft Teams, and answer from your own documents.
- Teams and telephony chatbots
- Research agents that draft the morning market brief
- LLM features inside existing products
OCR and entity-extraction pipelines that turn contracts, claims and clinical records into structured data you can query.
- Vision OCR and document NER
- High-accuracy entity extraction
- Search and retrieval over the extracted data
Churn, claim-denial, demand and cash forecasts, built end to end and monitored once they are live.
- Deposit and credit-card churn
- RFM and clustering segmentation
- Time-series forecasting for cash and demand
Research tooling for rates, FX and cross-asset desks, built on Bloomberg BQL/BQuant and Macrobond.
- BQL/BQuant and Macrobond data apps
- Backtesting and scenario analysis
- Dash apps for analysts and portfolio managers
Airflow pipelines with validation checks, feeding the Dash and Superset dashboards that executives read.
- Apache Airflow ETL with validation
- Plotly Dash and Apache Superset
- One store for research and reporting data
Getting models into production and keeping them there: deployment, a model registry, monitoring, CI and runbooks.
- REST deployment and model registries
- Monitoring, alerting and CI checks
- Runbooks and handover to your team
For investors
Technical due diligence on AI startups
Venture investors bring me in before a seed or Series A round to check whether a startup’s AI holds up, and afterwards to help the companies they back build it properly.
What the review covers
- Model architectureWhere the intelligence actually lives, what it costs to serve, and how hard it would be to copy.
- Data advantageWhether the proprietary-data claim holds up: rights, governance, and whether it improves with usage.
- InfrastructureDeployment, monitoring and MLOps maturity, with specific fixes for after the round.
- AgentsHow the agents behave when a tool fails, an input is malformed or a task runs long.
You get a written memo for the investment committee, and I’m available to talk it through.
After the round
- Architecture reviews on LangGraph, CrewAI and AutoGen
- Workshops and one-to-one sessions for engineering teams
- Retrieval tools that turn a data room into a first-draft memo
How I work
Four steps, one fixed price
- 1
Scope
An hour on the problem, the data and what “done” means. You get a written scope and a fixed price.
- 2
Build
Short sprints. You see working software every week and can change direction early.
- 3
Deploy
The system runs in your environment, on your infrastructure, with monitoring from the first day.
- 4
Hand over
You get the code, documentation and runbooks, and I walk your team through running it.
Questions about pricing, NDAs or working remotely? Read the FAQ
About
You work with me directly
I’m Qamar Zia, and QZ Solutions is my independent practice. My work has always been about getting machine learning out of notebooks and into systems people rely on.
I started in healthcare, where I owned a claim-denial prediction system from data ingestion to a monitored API. At TenX I led a team of four data scientists building churn, segmentation and cash-forecasting systems for retail banks. For a sovereign wealth fund I built BQL/BQuant data apps and backtesting tools for its bonds and FX research.
Since 2024 I’ve also worked on the investor side, running technical due diligence on early-stage AI startups at ReportEQ and Highwater Venture Partners. On every engagement, you work with me directly.
Based in Islamabad. Clients in the UAE, the US and Pakistan.
LinkedIn profileCareer
| 2025–now | Highwater Venture Partners | Senior AI Consultant: due diligence and portfolio advisory |
|---|---|---|
| 2024–2025 | ReportEQ | Senior AI Consultant: due diligence and agent evaluation |
| Consulting | Sovereign wealth fund | BQL/BQuant data apps, Dash analytics, backtesting |
| 2022–2024 | TenX | Data Scientist, leading a team of four on banking analytics |
| 2020–2021 | Cube Healthcare Systems | Associate Data Scientist: claim-denial prediction |
| 2015–2019 | COMSATS University Islamabad | BS Computer Science |
Certifications
- IBM Professional Data Scientist
- Microsoft Azure Data Scientist (DP-100)
- IBM Data Science Specialist (Level 2)
- 20+ IBM and Microsoft badges: data science, GenAI, Azure ML, watsonx, MLOps
Tools I use
Python, scikit-learn, TensorFlow, OpenAI, Anthropic and Azure LLM APIs, LangGraph, CrewAI, AutoGen, RAG and vector databases, Azure ML, IBM watsonx, Dataiku, Apache Airflow, Plotly Dash, Apache Superset, Bloomberg BQL/BQuant, Macrobond.
Contact
Start with a 30-minute call
Tell me what your team does by hand today, or which AI startup you’re evaluating. I’ll tell you honestly whether I can help, and I reply within one business day.