Building systems powering personalization and recommendations on Disney+.

Karat created the world’s first Interviewing Cloud to make the technical interviewing process fair, equitable, and enjoyable. Brilliant Black Minds, a free program by Karat, exists to increase the number of Black software engineers in tech. As an ambassador for the program, I support a community of over 8,000 Black software engineers as they practice for technical interviews and unlock access to new career opportunities.

◦ Designed and launched AI based grading system utilizing OpenAI LLM API, Python, and Weaviate (open-source vector database), allowing a tenfold expansion in terms of company’s reach.

◦ Architected and built a new notification system using AWS SNS and SQS, reducing notification times by 5x and lowering service costs by 90%.

◦ Developed proof-of-concept grading system using Python + DynamoDB to validate storage and management of evaluation criteria, results, and student submissions, serving as a foundation for a larger-scale build process.

◦ Achieved 100% job placement for participants in curated data structures and algorithms/SQL curriculum, equipping trainees with the technical skills necessary to succeed in software engineering job interviews.

◦ Led 30+ technical sessions on topics such as Python, Git, Docker, CML, DVC, MLflow, Airflow, Kafka, CI/CD, Testing, Dapps, and Smart Contracts, resulting in improved technical skills and knowledge for over 150 trainees.

◦ Consolidated and migrated relational data from 6+ sources spanning over 10 years, leading to enhanced data quality and enabling more effective data-driven decision making.

◦ Managed AWS access, deployed Redash dashboards, and implemented data quality checks using open-source tools like Great Expectations, empowering stakeholders with data insights and boosting confidence in decision-making.

◦ Built a layered, scalable ELT data engineering pipeline using Kedro, complete with transparent data lineage graphs for streamlined auditability and data processing reproducibility.

◦ Built core services for trend prediction, using various technical indicators, classification, regression, and clustering models. Wrote Smart Contracts for executing arbitrage opportunities – flash loans and cross-exchange arbitrage.

◦ Deployed Kubernetes-based infrastructure for trading bots, significantly improving scalability and ensuring seamless management of trade executions.

◦ Utilized Trino to process large-scale backtest datasets. Calculated and analyzed key financial metrics, including returns, Sharpe ratios, and drawdowns, providing valuable insights for decision-making and risk management.

◦ Transitioned to a serverless architecture, leveraging AWS Lambda and other cloud-native services, resulting in reduced cloud expenses of over 50%.

◦ Containerized Data Science playground notebooks using AuroraDB, Docker, MLFlow and Amazon S3. Dealt with automation of various system components using Airflow, AWS Lambda, and EventBridge.

◦ Managed end-to-end deployment, including DNS configuration on Route 53 and secure ECS deployment, for an automated React web app using GitHub Actions.