Mansour Shams

Accomplished data science and machine learning leader with expertise in data management, partner management, and fintech. Proven ability to enable cross-functional collaboration with Independent Software Vendors (ISVs), HyperScalers, and system integrators to deliver scalable AI/ML solutions. Skilled in fostering strategic partnerships, aligning business and technical goals, and driving mutual success within global ecosystems. Proficient in Python, C++, Go, TensorFlow, PyTorch, and cloud platforms (AWS, GCP, Azure). Demonstrated success in building scalable systems with RESTful APIs, Flask, Terraform, and CI/CD pipelines, leveraging cloud computing paradigms including SaaS, PaaS, and IaaS to enhance scalability and efficiency. Experienced in LLM fine-tuning, NLP, quantitative financial analytics, real-time fraud detection, and edge computing. Adept at managing data science teams to deliver reproducible and scalable models. Holds advanced degrees in Computer Engineering and Business, along with Lean Six Sigma Master Black Belt certification.

- Enhanced customer satisfaction by driving product features through storytelling within Scrum/Agile frameworks. Led backend architecture with IAM, EC2, Serverless Lambda, DynamoDB database design, Terraform (AWS, GCP, Azure) in .tf and .tfpl formats, and RESTful APIs.

- Leveraged the NICE cloud contact center platform to analyze transcribed customer interactions, implementing ML and Bedrock LLM models within AWS to identify call drivers, detect anomalies such as agent misbehavior or fraud, and deliver real-time, AI-driven insights that enhanced operational efficiency and compliance.

- Conducted novel research on contemporary AI topics, including training generative AI models (LLMs and text-to-image models) using AWS Bedrock, improving pre-trained models, and evaluating model performance. Experienced in working with large-scale LLMs with tens to hundreds of billions of parameters on AWS Bedrock.
- Integrated SaaS and PaaS solutions to streamline deployment of financial applications, enhancing system reliability and data management.
- Architected international microservices-based designs for real-time financial analytics applications on multi-cloud platforms.
- Deployed edge AI models for financial risk reduction on IoT devices.
- Established CI/CD pipelines using GitHub Actions and Octopus, integrating Docker and Kanban methodologies in a multi-threaded AWS environment. Utilized Postman for API testing and participated in the deployment and testing phases.
- Leveraged AI-assisted coding tools (Copilot, Intellicode, Tabnine, Snyk Deepcode) to enhance productivity, security, and code quality in Python, C++, Go, JavaScript, Snowflake, Kafka, and Flink for high-performance processing.
- Integrated NLP and Conversational AI models to reduce call costs and enhance fraud detection with CUDA.
- Pioneered innovative use cases in generative AI and NLU to understand call drivers and mitigate associated risks, enhancing operational efficiency and reducing potential financial losses.

PhD Computer Engineering