Open to conversations
FOUNDER — DOSSIER № 01
FileAktivcognit / Founder
DisciplineData Engineering & Analytics
Live time
KOB—
BasedKoblenz, Germany

Usama Khalid
Mirza↘
data, quietly engineered.

Founder & principal engineer of Aktivcognit. Six years architecting data platforms, streaming pipelines and analytical infrastructure across energy systems, fintech and healthcare — currently building data infrastructure for AI-powered HVAC optimisation.

Get in touch
Python Apache Spark Kafka Airflow PostgreSQL TimescaleDB MQTT Streaming Cloud Cost Optimisation Python Apache Spark Kafka Airflow PostgreSQL TimescaleDB Cloud Cost Optimisation
Section 01 — The founder
№ 01 / Founder UKM
U·M
Usama K. Mirza MMXX — MMXXVI

A data engineer by craft and a systems thinker by temperament — architecting and implementing scalable data platforms, streaming architectures, and ETL/ELT pipelines across energy, fintech, and healthcare.

Currently building data infrastructure for AI-powered HVAC optimisation: processing large-scale IoT and time-series datasets, designing analytical models, and putting automated data quality frameworks under production AI systems.

Practising AI-first engineering with Cursor and MCP tooling. Comfortable inside Python, Apache Spark, Kafka, Airflow, PostgreSQL, TimescaleDB, Docker — and the boring infrastructure that makes them actually ship.

Equally interested in cloud architecture and cost: right-sizing workloads, picking storage tiers that match access patterns, pruning the queries and jobs that quietly drain budgets, and treating the monthly bill as a first-class engineering signal.

RoleFounder · Data Engineer
BasedKoblenz, Germany
Emailusamak.mirza@gmail.com
Phone+49 152 1264 5328
Section 02 — A few numbers

Six years inside data — across three industries, four companies, and a few stubborn migrations.

Years in data
6yrs

Engineering pipelines, platforms and analytical infrastructure since 2020.

Sectors served
3·

Energy systems, fintech, and healthcare — each with its own data discipline.

Payment gateways
6+

Tabby, Tamara, Payfort, SADAD, Apple Pay & cards — wired into transactional reporting.

Budget leak found
13%

Identified through automated commission reconciliation and financial reporting on a fintech engagement.

Section 03 — Trajectory

A working archive — the roles, the rooms, the systems that taught the craft.

Oct 2025 —Present
Data Engineer
Energy / Building Intelligence
  • Scalable, cloud-native data platforms supporting AI-powered HVAC optimisation and building energy management.
  • ETL/ELT pipelines in Python and Apache Spark for large-volume IoT sensor, telemetry & energy data.
  • Real-time streaming via Apache Kafka and MQTT for HVAC and building automation ingestion.
  • Data quality and anomaly detection workflows feeding production AI reliability.
  • Migration of large-scale time-series workloads from ScyllaDB to PostgreSQL and TimescaleDB — cutting storage footprint and recurring compute spend in the process.
  • Workload right-sizing, query optimisation and storage-tier tuning to keep infrastructure cost proportional to value delivered.
GermanyRemote / On-site
Sep 2024 —Sep 2025
Working Student Data Engineer
Energy / Building Intelligence
  • ETL pipelines for HVAC sensor data, telemetry streams and energy consumption datasets.
  • Spark-based transformation jobs and automated ingestion workflows.
  • Kafka pipelines supporting near real-time analytics.
  • Warehouse modernisation across ScyllaDB, PostgreSQL and TimescaleDB.
  • Monitoring and validation tools improving data quality.
GermanyRemote / On-site
Dec 2023 —Jul 2024
Working Student Data Engineer
Fintech / SME Lending
  • Financial data platform operations through an organisational transition and acquisition.
  • ETL workflows processing loan, customer and transactional datasets.
  • Ingestion pipelines using Apache NiFi, Python, REST APIs and SQL transformations.
  • Python-based REST APIs for monitoring, reporting and risk assessment.
  • Validation, reconciliation and consistency checks on business-critical financial data.
GermanyRemote / On-site
Sep 2020 —Sep 2023
Intermediate Software Engineer
Healthcare & Fintech Engagements
  • Backend systems, integrations and reporting for healthcare and fintech applications.
  • Transactional data integration from 6+ gateways: Tabby, Tamara, Payfort, SADAD, Apple Pay, cards.
  • Analytical reporting on revenue, commissions and installment adoption.
  • Healthcare analytics dashboards tracking onboarding, patient growth and treatment adoption.
  • Identified 13% budget leakage via automated commission reconciliation.
RemoteMulti-region
Section 04 — The stack

Tools held long enough to be opinionated about — Python at the core, infrastructure all around it.

01 — Languages
Programming
Python SQL Java Ruby JavaScript
02 — Pipelines
Data Engineering
Apache Spark Kafka Airflow Pandas ETL / ELT Warehousing Modeling
03 — Sensors
IoT & Time-Series
MQTT HVAC Telemetry Sensor Pipelines Time-Series Analytics Energy Optimisation
04 — Storage
Databases
PostgreSQL TimescaleDB ScyllaDB Cassandra Redis
05 — APIs
Backend
FastAPI Flask Rails REST
06 — Ops
DevOps & Infra
Docker Kubernetes Grafana Nomad GitLab CI/CD CircleCI Git
07 — AI
AI & Automation
Cursor MCP n8n LLM-assisted dev Workflow Automation
08 — Quality
Testing & Monitoring
Great Expectations Unit Testing Data Validation Pipeline Monitoring Perf. Optimisation
09 — Cloud
Cloud Platforms
AWS GCP Azure Cloud-Native Serverless IaC Terraform
10 — FinOps
Cost Optimisation
Workload Right-Sizing Storage Tiering Query Optimisation Spot / Reserved FinOps Budget Alerts
11 — Realtime
Streaming Systems
Kafka MQTT Spark Streaming Event-Driven CDC Backpressure
12 — Signal
Observability
Grafana Prometheus Metrics Tracing SLOs Alerting
Section 05 — How I work

Pipelines are products. Data quality is not a dashboard — it is a nice-to-have contract. And the cloud bill is a design document — every line tells you which decisions paid off and which didn't.

i.

Built deliberately

Architectures sized to the problem, instrumented before they're scaled, and documented enough to survive the next teammate.

ii.

Maintained kindly

Boring is a feature. Migrations are planned around the people running them — and the pager that catches them at 3am.

iii.

AI-first, not AI-only

Cursor, MCP and LLM-assisted workflows as part of the discipline — used to ship more carefully, not less.

Section 06 — Credentials

A degree in progress, a degree completed, and three working languages.

Education

M.Sc. Web & Data Science
University of Koblenz · Germany · Oct 2023 — Present

Focus: machine learning, data science, big data analytics, distributed systems, enterprise applications. Currently selecting a thesis topic.

B.Sc. Information Technology
Quaid-i-Azam University · Islamabad · Sep 2016 — Dec 2020

Foundation in software engineering, databases, networking and applied computing.

Languages

English
C1
Urdu
Native
German
A1

Certifications

IELTS — English C1
Certified academic proficiency

Working language across all professional engagements.

Section 07 — Get in touch

Start a conversation.

Phone
+49 152 1264 5328
Based
Koblenz, Germany — working remotely with teams across Europe and further afield.
Availability
Open to selective conversations around data infrastructure, streaming architectures and analytics platforms.
Studio
Back to Aktivcognit ↗