Venu Gopal Reddy Janga

Facilitated the transition of the team to work on cloud services, providing hands-on training in AWS serverless services (Lambda, ECS). Created Sage Maker environments for the team and automated processes using Terraform templates. Focused on LLM deployment using Sage Maker jumpstart and tuned the llama2 for internal POC. Certified in prompt engineering and help many teams to kick start using GenAI. Working on the LLM serving with low cost and fine-tuning for specific project, and also exploring RAG techniques.

Advance knowledge of ML algorithms such as Ensemble Models, Ada-boost, Gradient-Boosting, Kernel-KNN, Grid search, Random Forest, Decision Tree, SVM, Linear Regression, Ridge, Lasso, Pipeline, Cross Validation, K-Fold and Extensive knowledge and implementation experience in XG- Boost algorithm Hyper- parameter tuning to improve the model, Validation metric for ML algorithms including classification report. In-depth expertise with a rich repertoire of Regression, Classification, Clustering and Dimensionality reduction algorithms. Lead deployment tool development for ML model deployment in AWS Sage Maker and Lambda, focusing on scalability and automation. Collaborated with the Data Lake team to orchestrate data pipelines with Prefect using AWS EKS services.

Experience in Data Analysis, visualization, statistics, programming, and Machine Learning techniques. Extensive Model building experience with Machine Learning algorithms for Product. Experience in creating mature Data science pipelines encompassing Data standardization, Feature extraction, model validation and optimization Exploring and visualizing data to drive insights Importing data from various sources and using numerous APIs. Designed CI/CD templates for all Fundamentals projects using GitHub Actions, migrating from Jenkins. Introduced Dockerization concepts and created data ingestion pipelines using AWS Glue, batch jobs, and Lambda. Implemented MLFlow using AWS ECS and S3 for experiment result logging. Designed and deployed scalable and cost-optimized solutions for all Fundamentals projects, resulting in a 70% cost reduction. Automated the ETF project, incorporating AWS Lambda for data ingestion, triggering training pipelines, and deploying 20 different models for 20 target columns in single API.

Guntur, India