AI/ML Engineer / Architect

Abdul Moiz

Ideas, built with intention.

Hands reaching, Creation of Adam in deep crimson

THE PHILOSOPHY

“I believe real AI engineering begins where cloud APIs end. Anyone can wrap a prompt, but true architecture is proven in the unseen: bias and variance rigor, data drift resilience, production evaluation, and knowing precisely when a human must guide the machine. I do not build demos that hallucinate. I build systems that endure.”

The Philosophy

What I help
you to ship...

Tools I use

Python
PyTorch
scikit-learn
OpenCV
FastAPI
Docker
PostgreSQL
Google Cloud

Machine Learning Systems

Models engineered for rigorous evaluation, observability, deployment, and ongoing monitoring.

MLOpsModel evaluationObservabilityMonitoring

AI Applications

Useful AI workflows with context, guardrails, and deliberate human handoff.

RAG systemsEvaluation loopsHuman-in-the-loop

Vision Intelligence

Computer vision pipelines that turn visual input into dependable signals.

Image analysisDetection pipelinesData quality

Production Architecture

Data, APIs, and infrastructure shaped for reliable operation as needs grow.

FastAPIContainersCloud deployment

Diffwise / Evidence-bound PR review

Code reviews grounded in evidence, with humans in control

In production

Locked

Diffwise

Pull-request review built on Trailforge

Private case study

Agentic review
Human approval
Durable workflows

Trailforge / Custom agent harness

Bound the work. Preserve the trail. Keep authority with the host.

View project ↗

Trailforge

Provider-independent agent execution with durable recovery

Agent harness
Evidence & verification
Durable workflows

StockPredictor / Daily equity forecasting

Forecast the move, rank the week, and log every call

View project ↗

StockPredictor

Calibrated move-size forecasts and a five-session ranking, scored live

Time-series ML
Walk-forward evaluation
Live retraining

HandoffCheck / Product to engineering

From a feature request to a brief both teams agree on

View project ↗

HandoffCheck

Requirement review, shared approvals, and implementation briefs

AI assistance
Two-lead approval
Team workflows

Showing 4 of 7 projects

Kinetiq Solutions
Promoted
AI/ML Team Lead
Jul 2026 - Oct 2026Remote
Leading a cross-functional AI/ML team, setting the technical direction for GenAI and machine-learning initiatives while mentoring engineers across model design, training, and deployment.
- Leadership
- GenAI / ML
- Model deployment
Kinetiq Solutions
AI/ML Engineer
Sep 2025 - Jun 2026Remote
Built production ML and LLM systems end to end, from data preparation and evaluation to FastAPI integration, vector stores, observability, and post-deployment monitoring.
- MLOps
- Model evaluation
- Observability
- FastAPI
Digifloat
Artificial Intelligence Intern
Apr 2026 - Jul 2026Hybrid
Developed real-time computer-vision and robotics perception systems, taking image data from preprocessing through low-latency object detection and tracking.
- ROS2
- YOLO
- Gazebo
- Tracking
Prosilient Systems Inc.
AI Researcher
Jul 2025 - Sep 2025Remote
Engineered automated research-data collection pipelines and transformed unstructured web data into cleaner, more dependable inputs for downstream model training.
- Web scraping
- Data pipelines
- Data quality
CARE
Center for Advanced Research in Engineering
AI Developer
Jul 2024 - Sep 2024On-site
Built RAG document-intelligence and NLP systems, connecting models to FastAPI and Flask services with a React interface and a production-minded delivery workflow.
- RAG
- LangChain
- FastAPI
- React