Hi, I'm Aman Khokhar.

Machine Learning Engineer

View My Work

About Me

I am a passionate Machine Learning Engineer currently based in London, United Kingdon with experience in building, deploying, and scaling data-driven solutions. I am experienced in MLOps, with a strong understanding of automating ML workflows, model deployment, and lifecycle management, and skilled in implementing CI/CD pipelines to ensure seamless integration, continuous monitoring, and scalable deployment of machine learning models across diverse environments. My areas of expertise include Large Language Models (LLMs), Natural Language Processing (NLP), and Generative AI.

Skills

Programming Languages

Deep Learning and Computer Vision

Data Visualisation

LLMs & Natural Language Processing

Database management

Cloud & MLOps

Career & Education

Machine Learning Engineer (Freelance)

ECAD Architectural Startup

Jan 2025 - March 2025

Developed a computer vision pipeline combining OpenCV and photogrammetry to create elevation plans for UK houses from video data

M.Sc. Computer Science

Brunel University London

2023 - 2024

Specialized in AI with a focus on deep learning architectures and NLP

Research Assistant

Delhi Techonological University (DCE)

2021 - 2022

Designed an end-to-end computer vision pipeline in Databricks to automate detection of damage on solar panels, including instance segmentation and binary classification of thermal drone imagery.

M.Tech. in Power System Engineering

Delhi Techonological University (DCE)

2020 - 2022

Specialized in optimising power systems using machine learning technologies

Project Engineer

Cable Corporation of India Ltd.

2019 - 2020

Deployed electrical load forecasting models using Amazon Forecast, achieving a 20% improvement in forecast accuracy.

B.Tech. in Electrical & Electronics Engineering

Delhi Techonological University (DCE)

2014 - 2018

Built a solid foundation in algorithms, data structures, and problem-solving.

Projects

LLM Model

Custom Transformer Language Model

Built a transformer language model with 2 decoder layers, 256-d embeddings, 4 attention heads, and a custom BPE tokenizer (5,000 tokens), totaling 3.62M parameters.

Github
Hand Gesture Controller Demo

Train Big, Tune Tiny: A Practical Guide to LoRA-Based Fine-Tuning of LLMs

Medium Article on Fine-Tuning Large Language Models

Medium Article
XRay Image 1 XRay Image 2 XRay Image 3

Chest X-Ray Diagnoses model

Built an EfficientNet-based image classification model for diagnosing chest X-rays for cardiomegaly. Analyzed and reported model performance using confusion matrices, achieving a high recall of 0.87 with an accuracy of 83.5%.

Github
ChatBot GIF

AWS Documentation Chatbot with RAG

Developed an AI-powered chatbot using Python, LangChain, and OpenAI API, processing 10,000+ AWS documentation pages into markdown files and chunking them for efficient retrieval. Implemented OpenAI embeddings to vectorize 66,000+ chunks in ChromaDB for fast, similarity-based searches. Built an interactive Gradio-based chatbot interface.

Github
Audio Project Image 1 Audio Project Image 1

Conditional Audio Latent diffusion model

Developed a Stable Diffusion-based text-to-image model using Mel spectrograms from the AudioCaps dataset (46K samples). Integrated CLAP for text-to-audio embeddings and HiFi-GAN for phase recovery, achieving high-fidelity (up to 90dB) audio reconstruction.

Github
Hand Gesture Controller Demo

Hand Gesture Controller

Developed a system that controls the mouse pointer using hand gestures. The project leverages real-time computer vision techniques and machine learning to detect and track hand movements.

Github

Certifications

AWS Certified Machine Learning Engineer - Associate

AWS Certified Machine Learning Engineer - Associate

Amazon Web Services

2024/12/30 - 2027/12/30 View Certification
IBM Data Science Professional Certificate.png

IBM Data Science Professional Certificate

IBM

2025 View Certification

Contact Me

If you have any questions or opportunities, feel free to reach out.