AR / Portfolio

Ali Raza

Shipping production ML systems across vision, voice, and agentic AI.

2

Years Experience

19

Projects Shipped

2

Publications

5

Domains Covered

Selected work

Featured Projects

Toolbox

Skills

Machine Learning & Deep Learning

PyTorchTensorFlowKerasScikit-learnHuggingFace TransformersLoRA / QLoRA / PEFTXGBoostLightGBMFine-tuningTransfer Learning

Computer Vision & Medical Imaging

YOLOv8TransU-NetU-NetViTSAMResNetDenseNetBioViL-TGradCAMCBAMOpenCV

Generative Models

GANsVAEsDiffusion ModelsFlow MatchingKnowledge DistillationMamba (SSM)

NLP, LLMs & RAG

BERTLangChainLangGraphLlamaIndexRAG PipelinesFAISSPineconePrompt EngineeringOpenAI APIGemini APIAnthropic API

MLOps & Deployment

MLflowDockerKubernetesKubeflowPrometheusGrafanaEvidently AIFastAPICI/CDGitHub ActionsAWS EC2Vast.ai

Data & Languages

PythonC++JavaScriptTypeScriptSQLPandasNumPyPostgreSQLMongoDBGit

Frontend & Tools

ReactNext.jsVitePower BINginx

Career

Experience

Machine Learning Engineer

HusQuay Solutions · Remote, via Crew Innovations, Faisalabad

Dec 2024 - Present
  • Built and shipped computer vision and ML features for a production property-tech SaaS platform serving US customers.
  • Diagnosed the main accuracy bottleneck in an image-understanding pipeline and redesigned it into a multi-stage detection, segmentation and vision-language model pipeline, improving fine-grained classification with zero hallucinated outputs in early validation.
  • Replaced a planned GPU retraining cycle with an API-driven inference approach, cutting iteration time and compute cost.
  • Owned the ML pipeline end to end, from raw image upload to automated, customer-ready output in a live production system.
  • Shipped via FastAPI and Docker with MLflow model versioning and GitHub Actions CI/CD, working closely with product and engineering teams.

Full Stack Engineer

Freelance (Self-employed) · Faisalabad, Pakistan · Remote

June 2024 - Present
  • Built and deployed multiple full-stack websites end to end (design, frontend, CMS integration, deployment) for local and international businesses, including content-heavy platforms with searchable catalogues and admin-managed content.
  • Delivered data analysis projects for businesses: scraped and cleaned market data, mapped competitors and demand by area, and turned the results into recommendations on underserved markets.
  • Took each project through structured client feedback rounds, from first prototype to final handover.

Data Analyst Intern

TAKMIL · Louisville, Kentucky, USA (Remote)

2024
  • Cleaned and analyzed large-scale datasets via Python (Pandas, NumPy, SQL) and built Power BI dashboards.
  • Proposed optimizations that improved workflow efficiency by 30%.

About

A short version

Photo of Ali Raza

I'm a final-year BSc Computer Science student at FAST-NUCES Faisalabad (graduating 2026) and a freelance engineer since 2023, building production ML systems and full-stack web products for local and international clients, across computer vision, medical AI, NLP, voice, and agentic workflows, end to end from data and training through deployment and monitoring.

My work spans two registers: research rigor, multimodal error detection in AI-generated radiology reports, GAN-based disease progression modeling, a co-authored hybrid transformer and CNN abstract accepted at ACR 2026, and a training-free lung CT registration method accepted at Learn2Reg 2026 (MICCAI) with an invited oral presentation, and production engineering: FastAPI services, Kubernetes deployments, MLOps pipelines, and RAG and voice agents shipped for real users. I care about systems that are useful, measurable, and honest about where they fail.

Education

BSc Computer Science

FAST-NUCES, Faisalabad

2022 - 2026

Thesis: RadGuard, Multimodal Error Detection in AI-Generated Radiology Reports.

Publications

Raza A. “Inter-Lobe Boundary Channels and Localised Folding Repair for Expiration to Inspiration Lung CT Registration.”

Learn2Reg 2026 Challenge, MICCAI 2026, Strasbourg. Invited oral presentation and poster.

Accepted

Service & Activities

  • Co-Head, Hackathon, ACM Society, Daira 2025
  • Member, ACM Society Management Team, Daira 2024

Contact

Let's talk

Open to AI/ML engineering roles, research collaborations, and PhD opportunities.