London Area, United Kingdom
Engineer with 3+ years at Amazon and Elton, combining software engineering expertise with AI and machine learning specialization. Experienced in developing and optimizing AI models using cloud infrastructure and AWS services. Proven track record in architecting scalable AI solutions, particularly in NLP and recommendation systems. Led projects integrating AI capabilities into existing platforms in Amazon Connect, enhancing user experiences while maintaining system stability.
Spearheaded development of NLP pipeline using transformer models and PyTorch, improving text classification accuracy by 35% for customer support automation. Architected scalable machine learning platform with Python, Django, and MLflow, enabling efficient model versioning and deployments. Implemented LLM fine-tuning framework, improving domain-specific task performance by 28%. Developed and optimized deep learning models with PyTorch for image recognition, achieving a 15% accuracy boost and 30% faster inference. Improved financial document parsing accuracy by 15% using Python/TensorFlow NLP models. Deployed ML inference endpoints on Google Cloud Platform for scalable customer behavior analysis. Reduced API response latency by 30% through async I/O optimization and query batching in Django REST. Led integration of vector database with MySQL for similarity-based recommendation systems, improving query speed by 60%. Conducted A/B testing and statistical analysis, resulting in a 25% improvement in system efficiency. Reduced production issues by 44% by implementing CI/CD pipelines with Jenkins and GitHub Actions for ML deployments.
Organized and hosted weekly technical workshops on Python, web development, and introductory AI/ML concepts for 50+ student members. Led ACM Hackathon planning committee, coordinating with sponsors, mentors, and university departments to host events with 150+ participants. Introduced a “Code Sprint” competitive programming series, boosting club engagement by 40% and attracting participants from across the Five College Consortium. Managed ACM’s GitHub repositories and project collaborations, implementing structured workflows and mentorship for new contributors. Partnered with faculty to run career development sessions, including resume reviews, mock interviews, and industry speaker events.
Optimized machine learning models for Amazon Connect's natural language processing pipeline, improving customer intent recognition accuracy by 15%. Reduced analytics infrastructure costs by $200K/year through compute optimization and intelligent auto-scaling across Amazon Connect services. Re-architected pipeline processing over 100M events/month, reducing ML feature engineering time by 40% and enabling real-time sentiment analysis. Leveraged AWS services (SageMaker, Lambda, EC2, DynamoDB) to deploy scalable AI solutions for real-time speech analytics and sentiment detection. Streamlined feature engineering workflows, optimizing data schemas and reducing data delivery latency by ~40%, accelerating ML experimentation cycles. Achieved 99.99% uptime SLA by implementing health checks and automated failover mechanisms. Built real-time analytics dashboards serving 20+ internal teams, reducing manual reporting time by 60%. Reduced query response times by 35% through DynamoDB partition key optimization and intelligent caching strategies. Led integration of AI capabilities into cloud infrastructure, improving performance and reducing latency in language processing tasks.
Developed internal tools and scripts in Python to automate grading and data validation processes for Computer Science courses. Created a lightweight web-based dashboard for tracking student performance and assignment completion using Flask and SQLite. Assisted in debugging and optimizing course-related applications, improving load times by 25%. Collaborated with faculty and student assistants to streamline workflow, reducing administrative overhead by 20%. Managed version control with Git, ensuring smooth collaboration across multiple contributors.