Akash Dixit

Leading, mentoring and nurturing a global team of data scientists, data engineers and analysts. Strategizing go-to-market (GTM) efforts to research, build & scale current AI offerings. Collaborating with McKinsey’s Partners to identify strategic priorities and build clear, actionable project roadmaps. Leveraging AWS services, Kubernetes, and tools like MLFlow to build, automate and scale workflows involving data pre-processing, feature engineering, CI, CD and automated retraining.

Leading a team of data scientists in India and collaborating with leadership to build in house capabilities focusing on natural language processing capabilities in speech, text and vision. Acting as a product owner and translating business problems into key initiatives. Productizing machine learning models with low latency via containerizing, AWS ECR, and Fargate/EKS instances. Code refactoring and building reusable pipelines using python, docker, FastAPI, CI/CD integration and AWS Cloud.

Acting as an individual contributor solving difficult problems in workforce and HR domain. Collaborating with Country managers to work on a roadmap leveraging data science and analytics to build intelligent solutions addressing key challenges.

Working in insurance and reinsurance industry with underwriters and catastrophe SMEs on data science solutions focused on Image classification leveraging deep learning models, recommendation, classification, and regression models. Actively engaged with CAT team SMEs, and Underwriters to address data issues and building a standardized code pipeline. Spearheaded development of a deep learning model for US region to detect early signs of catastrophe events.

Generating insights and dashboard for customers insurance policies by Tableau.

Enabling automation for manual tasks by using VBA, MS-Access. Creating reports and insights against billed customers. Risk and trend analysis for decline in the customers revenue across COLT products. Writing queries and extracting the data from SQL Server using T-SQL.