Looking for a job

Dhruv Bhavsar

@dhruvbhavsar1

AI engineer at Wappnet Systems who cut maritime routing p99 latency from 2 seconds to under 100 ms.

At Wappnet Systems, I optimized a maritime routing API for the Finnish government with Contraction Hierarchies, cutting p99 query latency from 2 seconds to under 100 ms. The MVP was accepted for patent filing.

I deployed pharmacy voice AI agents handling over 500 calls a day with under 50 ms average reply latency, and shipped a RAG-based front-desk widget that onboarded 20+ customers. I also architected osora.ai, an enterprise multi-agent copilot on AWS Bedrock that reduced LLM token consumption by 63% per task.

At ISRO, I built a TensorFlow/Keras LSTM+CNN model for solar irradiance forecasting that reached 79% accuracy on satellite data. I also built rustlette, an open-source Rust/Python drop-in for Starlette, and TeleAI Indic, a real-time voice AI platform for Indian languages.

Experience

Work history, roles, and key accomplishments

SDE 2, AI and Data Science

Wappnet Systems

Jan 2025 - Jul 2026 (1 year 6 months)

Led a team of 6-7 AI engineers, optimizing a maritime routing API for the Finnish government and deploying pharmacy voice AI agents. Shipped a RAG-based AI front-desk widget and owned system architecture and client delivery.

Contraction Hierarchies graph algorithms Voice AI RAG AWS System Architecture Team Leadership

AI Engineer

Wappnet Systems

Jul 2024 - Dec 2024 (5 months)

Architected an enterprise multi-agent copilot on AWS Bedrock, reducing LLM token consumption by 63%. Delivered a healthcare CRM search copilot with fine-tuned embedding models and RAG.

AWS Multi Agent Systems Redis PostgreSQL RAG Elasticsearch Fine Tuning GPU

Data Engineering Intern

Wappnet Systems

Jan 2024 - Jun 2024 (5 months)

Engineered 53 Apache Airflow DAGs for healthcare data ingestion and transformation, and deployed an auto-scaling geocoding service on AWS Lambda.

Airflow AWS Lambda Polars PySpark DBSCAN Data Engineering

Research Intern, Machine Learning

Indian Space Research Organisation (ISRO)

Oct 2023 - Jan 2024 (3 months)

Built an LSTM+CNN hybrid model for solar irradiance forecasting on satellite data, achieving 79% accuracy and outperforming previous models. Trained on multiple GPUs and set up continuous retraining.

TensorFlow Keras LSTM CNN CUDNN GPU Training Hyperparameter Tuning

Education

Degrees, certifications, and relevant coursework

University of California, Los Angeles

Master of Engineering, Artificial Intelligence

2026 -

Pursuing a Master of Engineering in Artificial Intelligence with expected graduation in September 2027. Coursework includes Automated Reasoning, Fundamentals of Artificial Intelligence, and Engineering 214.

Gujarat Technological University

Bachelor of Engineering, Computer Science and Engineering

2020 - 2024

Grade: 9.60/10

Earned a Bachelor of Engineering in Computer Science and Engineering with a focus on AI/ML, achieving a CGPA of 9.60/10. Relevant coursework included Operating Systems and Introduction to Java.