mohammad Ibrahim Khan

Architected and deployed a real-time AI proctoring platform on AWS EC2 supporting 100+ concurrent users
with sub-200ms latency using FastAPI and WebSockets, taking the project from prototype to production.
• Developed a multi-modal anomaly detection system combining computer vision (DeepFace) and NLP to
identify suspicious behavior, improving detection accuracy by ~40%.
• Worked with microservices-based architecture and containerized deployments using Docker and
Kubernetes
• Designed data-driven decision systems to monitor user activity and trigger automated actions based on
predefined thresholds.
• Built centralized violation tracking system using Zustand with configurable thresholds, enabling real-time
alerts and automated exam termination — reducing unfair exam behavior by ~60%
• Engineered browser integrity enforcement (tab-switch detection, fullscreen exit, DevTools monitoring) and
camera integrity checks (device locking, FPS/resolution monitoring), ensuring consistent and tamper-proof
exam sessions
• Built and deployed containerized microservices using Docker and Kubernetes with CI/CD pipelines via
GitHub Actions, ensuring reliable, maintainable production deployments.
• Optimized system performance using Redis caching and real-time monitoring with Grafana.

• Developed and deployed a dynamic dashboard using React.js, TypeScript, Tailwind
CSS, and Node.js to deliver real-time business insights.
• Implemented CI/CD automation with GitHub Actions and blue-green deployment,
reducing system downtime by over 90%.
• Configured Nginx and SSL certificates to enhance production reliability, scalability,
and security.
• Designed and integrated interactive data visualization cards to track key financial
metrics such as invoices, salary vs. billing, and recent payment trends.
•
Automated business workflows, including salary slip generation, expense approvals
(via MailerSend), and contract/ID expiry notifications to improve operational
efficiency.
• Developed secure REST APIs with Role-Based Access Control (RBAC) to ensure data
integrity and controlled access.
• Engineered a Retrieval-Augmented Generation (RAG) pipeline leveraging LangChain
and GPT models, improving chatbot response accuracy by 30%.
• Implemented advanced memory buffering and prompt-chaining techniques to
enable contextual awareness and maintain conversation continuity.
• Executed end-to-end data workflows, including data preprocessing, feature
engineering, and API testing using Postman for model validation.
• Optimized backend performance by leveraging edge functions, significantly reducing
latency and improving response times.
• Diagnosed and resolved production issues related to deployment pipelines,
environment configuration, and Hot Module Replacement (HMR), ensuring system
stability.
• Collaborated in agile sprints to design and implement automation solutions aligned
with real-world business requirements.

Worked on implementing machine learning models in real-world scenarios under
structured guidance.
Performed data preprocessing and feature engineering for model training and
evaluation.
Applied model optimization techniques to improve accuracy and reliability.
Gained practical experience in the complete ML development lifecycle.

Gained proficiency in Python libraries including NumPy, Pandas, Matplotlib,
and Scikit-Learn.
Applied algorithms like KNN, Support Vector Machines, and Random Forest to
solve real-world problems.
Conducted error analysis, optimized model performance using cross-validation
techniques, and implemented feature selection strategies.

• Developed and optimized machine learning models using KNN, SVM, and Random
Forest algorithms on real-world datasets to drive data-driven insights.
• Executed end-to-end projects, including Titanic Survival Prediction and Banking
Churn Prediction, achieving high model accuracy and reliability.
• Applied cross-validation, feature selection, and hyperparameter tuning to enhance
model performance and generalization.
• Utilized Pythons data science stack NumPy, Pandas, Matplotlib, and Scikit-learn for
data preprocessing, visualization, and model evaluation.