Vedant Jawanjal

Advisor Chat Transcript Platform: Built AWS + PySpark pipelines to ingest customer interaction data; engineered logic to merge chat events into unified conversations; designed a Kafka to Elasticsearch streaming pipeline handling 10M+ records/day; optimized Elasticsearch indexing; developed Flask REST APIs. GenAI Guardrails & Safety System: Strengthened safety systems for detecting jailbreaks and harmful intent; used Ollama for synthetic adversarial prompts; leveraged Hugging Face models for classification; performed embedding experiments; fine-tuned safety classifiers (XGBoost, Transformer models); curated large-scale safety datasets.

Enhanced and maintained scalable data pipelines and data warehouses using PySpark, Python, and AWS services (S3, Glue, EMR, Athena, Airflow). Collaborated with cross-functional teams to implement complex business requirements in an Agile environment. Developed complex SQL scripts for data analysis. Automated ETL workflows in Databricks and AWS Airflow. Conducted root cause analysis and performance tuning. Led migration of legacy ETL processes to Databricks. Spearheaded deployment of real-time data analytics for pharmaceutical clients. Implemented data validation and profiling techniques.

Supported 3+ web development projects using HTML, CSS, Python, and the Django framework.

Architected a Python-based data collection pipeline for IoT sensors, increasing data reliability from 70% to 98%.