Hamza Mehmood — AI Engineer

01Proof

Three hard problems. Working systems.

Implemented

The problem: Infrastructure gets sized once, then keeps paying for capacity the workload no longer needs.
The solution: Engineering workflows that weigh price, specs and regional availability together, then recommend a better-fit VM.
The outcome: [VERIFIED RESULT]

Interactive demo · sample data

Current

vCPU needed

Memory needed

Evaluation

price · specs · region

Better fit

vCPU needed

Memory needed

specs ✓

Simulated catalogue and prices for illustration — not customer data.

Contributed

The problem: Large energy datasets are only useful when people can question them quickly.
The solution: Data infrastructure and pipelines on PostgreSQL and TimescaleDB, served through backend services.

Interactive demo · sample data

Sources

Ingest

Backend services

validate · normalise · batch insert

TimescaleDB hypertable

Insight

points rendered

Load · kW (sample)

Query

Generated sample series, illustrative pipeline — not production data.

The outcome[VERIFIED RESULT]

The problem: Misconfigurations hide across accounts, services and regions until someone looks.
The solution: Cloud security analysis and scanning workflows, including hands-on evaluation of Prowler-based capabilities.
What's real: Prowler-based checksEvaluated

Scanning workflow[IMPLEMENTED / PROTOTYPE]

Findings dashboard (this demo)Concept
The outcome: [VERIFIED RESULT]