Sharif Ahamed

Languages

Advanced in Python. Familiar with C++, Rust.

ML & Deep Learning

PyTorch, Hugging Face, LLMs, RAG, Fine-tuning, Inference Optimization, Agentic AI.

Cloud & DevOps

AWS, Docker, CI/CD, Git.

Databases

PostgreSQL, MySQL, ElasticSearch, Vector Databases.

Amazon Fulfilment Centre

Team Lead, L3 • August 2024 — February 2026 • Leeds, UK (Full-Time)

• Did this just to make the money; but learned a ton regarding working in high paced environment where millions of orders are process.
• Improved my leadership skill a lot by building and managing my own team and at the same time maintaining Amazon standard of work in terms safety and efficiency.
• Led a team of 25 associates on Amazon's robotics floor, managing daily fulfillment operations and ensuring smooth coordination with automated systems.
• Oversaw shift execution and safety for 500+ associates around autonomous robotics system.
• Maintained KPIs related to system uptime, throughput, and floor efficiency.

Gigalogy Inc

Machine Learning Engineer • March 2020 — December 2022, May 2023 — August 2024 • Hybrid (Full-Time)

• Improved the multilingual and multi-modal data processing pipeline for a dynamic RAG-based chat system (Maira). Optimized dynamic prompt generation for 10+ products, significantly improving model response and adaptability across different tasks.
• Developed and deployed cost-effective BERT-based classifiers to filter approximately 30% to 50% of simple user requests (depending on clients), significantly decreasing dependency on resource-intensive large models through LLM routing.
• Led development of a local and cloud-based system for SFT-based fine-tuning of LLMs, integrating performance monitoring and chat support. Optimized time to first token from 900ms–1200ms to 400ms using continuous batching, prompt compression, and model quantization.
• Built the REST API-based core backend system for a dynamic personalization platform offering personalized search, image search, and dynamic pricing.

University of Bradford

Master's with Distinction in Artificial Intelligence • 2024

• Conducted master's dissertation research focused on solar XRAY-flux data preprocessing, fractal interpolation for data imputation, and fractal dimension for self-similarity analysis.
• Developed a real-time Streamlit-based predictive system that forecasts solar flares with classification.

North South University

Bachelor's in Computer Science and Engineering • 2019

• Completed comprehensive undergraduate program in Computer Science and Engineering.
• Built foundational knowledge in algorithms, data structures, software engineering, and computer systems.

Mercor Cheating Detection

20th place / 419 teams (2025)

Placed 20th out of 419 in a Kaggle competition focused on graph based fraud detection, emphasizing feature engineering over complex modeling. Built a tabular model based on added graph features, and used conservative pseudo-labeling with calibrated ensemble models for final submission.

Nivolytics - AI-Powered Investment Analytics

Personal Project (2025)

Built Nivolytics, an AI-powered investment analytics platform using FastAPI, Vue.js, and Perplexity AI's Sonar models for real-time and deep financial insights. Integrated LLM-based analysis for instant portfolio summaries, multi-source due diligence, and intelligent investment research.

Solar Flare Prediction System

Master's Dissertation (2024)

Conducted master's dissertation research focused on pre-processing solar XRAY-flux data, fractal interpolation for data imputation, fractal dimension for self-similarity analysis, and a real-time Streamlit-based predictive system that forecasts solar flares with classification.

Cohere Prompt Challenge

Winner (2023)

Won the Cohere Prompt Challenge. The task involved testing and challenging the Cohere Coral chat to identify vulnerabilities, such as generating misinformation, inducing hallucinations, and generating responses it is designed to avoid.