Sharad Jain

Led AI strategy implementation, developed OpenAI assistant-based Salesforce digital worker, streamlined admin processes, and optimized Gen AI performance. Built a synthetic data generation pipeline for schema-aware databases using LLaMA 3.1 405B model using fireworksAI, and generated smaller state-of-the-art models for data augmentation. Implemented multi-agent collaboration for LLM applications using OpenAI ASsistant API. Developed an alternate approach to multi-agent collaboration using Llama-stack of APIs to Technologies: OpenAI, LLaMA, Mistral, LangChain, LlamaIndex, FireworksAI

Led cross-platform marketing analytics and implemented multi-touchpoint attribution framework. Developed ETL pipelines to collect and unify user identities across mobile and desktop platforms, solving the challenge of single users accessing through multiple devices. Engineered complex SQL queries to stitch user identities and create a holistic view of user journeys. Built and maintained data pipelines using Airflow for automated data processing and analysis. Implemented advanced data quality checks to ensure accuracy in cross-device user tracking. Drove data-driven decision making by providing insights on user behavior across different touchpoints. Technologies: Python, Airflow, SQL, Looker, AWS, ETL, Identity Resolution

Implemented MLOps practices, engineered scalable data pipelines, optimized internal data engineer user onboarding, and led ML workflow migration to Kubernetes. Technologies: Apache Spark, Airflow, Kubernetes, A/B testing

Implemented machine learning models for infrastructure failure prediction and built business insights dashboards. Technologies: AWS QuickSight, Machine Learning

Designed ETL mappings, performed impact analysis, and managed data repositories. Technologies: IBM InfoSphere, ETL