HELLO, I'M
Yasir Javed Khan
Machine Learning Developer • AI Systems Builder
I work across data, model development, experimentation, evaluation, and deployment — building ML systems that are measured and documented, not just claimed.
Connect with me
HELLO, I'M
Machine Learning Developer • AI Systems Builder
I work across data, model development, experimentation, evaluation, and deployment — building ML systems that are measured and documented, not just claimed.
Connect with me
ABOUT
I build machine learning systems that turn data into usable software. My work focuses on Python, classical machine learning, deep learning, computer vision, NLP, and model deployment. I enjoy taking projects from experimentation to working applications.
From data exploration to deployed endpoints — structured, reproducible, and focused on working software.
Test hypotheses, compare models, and measure what matters.
Turn notebooks into maintainable systems with clean interfaces.
Validate on held-out data and ship to usable applications.
WHY I BUILD
I am interested in how data, models, and software combine to form usable systems. I build to test ideas, measure outcomes, and learn where models break — then iterate with clearer constraints and better data.
TECHNICAL STACK
Grouped by capability. Each area reflects hands-on use across projects — from data preparation to model development and deployment.
ML LAB
Hands-on exploration — comparing models, measuring trade-offs, and documenting findings without fabricated results. Each experiment follows build → evaluate → document.
Comparing training from scratch against pretrained backbones on image classification; tracking accuracy, training time, and generalization.
Computer Vision • PyTorchSystematic evaluation with train/validation/test splits, confusion matrices, precision/recall, and ROC analysis on held-out data.
Evaluation • Scikit-learnGrid and randomized search for learning rate, depth, regularization, and batch size — logging validation curves.
Tuning • ValidationExploring attention-based models for text, including tokenization, sequence handling, and fine-tuning workflows.
NLP • TransformersExploratory analysis of feature distributions, correlations, missing data, and preprocessing decisions before modeling.
EDA • PandasProfiling inference latency, model size trade-offs, and deployment readiness for practical use.
Optimization • DeploymentEvidence over adjectives — every stage is measured and documented.
FLAGSHIP PROJECT
MOST DETAILEDSELECTED WORK
Four featured systems spanning ML, deep learning, NLP, and computer vision — with full technical detail and inspectable demos. The complete catalog holds 15 builds.
TRACK RECORD
Verified competitions, certifications, and systems shipped — each tied to measurable work.
Secured 1st place in the interschool coding competition organized by CodeVerse.
Awarded for outstanding academic performance in Grade 10.
Completed the Python Programming course (Udemy) with hands-on project work.
Participated in an AI & ML workshop and built real-world mini projects.
Built and shipped 5+ end-to-end systems including ML prototypes and web applications.
May 2024GET IN TOUCH
Have a system to build or a model to evaluate? Reach out — I prioritize clear problem statements and working prototypes.
Connect with me
