I’m Atharva Kadam, an AI/ML Software Engineer with a strong background in building scalable computer vision and AI systems. I enjoy solving real-world problems using a combination of machine learning, distributed systems, and modern MLOps. Currently, I work at Droisys Inc., where I lead initiatives in brand detection, generative AI, and AI-powered analytics.
Software Engineer II @ Brellium
Feb 2026 – Present, New York, NY
- Building Brellium - an AI-powered clinical compliance platform
Data Engineer @ Droisys Inc
May 2025 – Feb 2026, Las Vegas, NV
- Engineered a high-performance, YOLO-based computer vision system by adopting a multi-model architecture and multi-GPU training strategy, increasing brand detection accuracy by 25% and reducing training time by 80%. Built and deployed GPU-accelerated inference using FastAPI and NVIDIA Triton, improving inference speed by 30% over previous Flask-based setups.
- Led the revamp of Data and MLOps infrastructure with multi-GPU data parallelism, automated retraining, and CI workflows—cutting training time by 60–70% and saving 10–12 hours of weekly ops overhead. Built full data pipelines covering scraping, preprocessing, annotation, versioning, deployment, and monitoring to enable scalable model lifecycle management.
- Reduced creative turnaround time for ad campaigns by 95% through scalable generative AI pipelines using diffuser-based inpainting and LoRA-finetuned LLMs, leveraging Stability AI frameworks for rapid deployment.
- Architected a LangGraph-based multi-agent AI system integrated with RAG pipelines (MongoDB Atlas vector search + structured filtering), enabling sales teams to extract insights from unstructured PDFs via natural language and visualize results in context (e.g., on maps), with sub-second response times.
- Designed a two-stage CV pipeline for brand/price detection using OCR and transformer-based string matching, boosting accuracy by 30%. Integrated the model into a FastAPI service for real-time shelf analytics, contributing to 10–15% sales growth for key alcohol distributor clients.
- More details in my resume or LinkedIn
Computer Programmer Analyst 1 @ Droisys Inc
May 2023 – April 2025, Las Vegas, NV
- Developing and implementing data engineering and machine learning techniques to solve business problems and drive customer value
- Using state-of-the-art machine learning models and neural networks for solving object detection problems
- Employing Python to develop data pipelines that efficiently assist in the building, training, and scaling machine learning and AI solutions for business applications
- Implementing best practices for data modeling, data quality, and data governance
- Employing Python and Flask to develop efficient and scalable machine learning and AI solutions for business applications
MTS Software Engineer II @ NetApp Inc
June 2022 – April 2023, San Jose, CA
- Led development of a scalable Audit Logging service using .NET, Java, and MySQL.
- Resolved key bugs and improved customer experience for major product releases.
Languages:
Python, Java, C, C#, .NET, Ruby, TypeScript, Bash
AI/ML & Data:
PyTorch, TensorFlow, Scikit-Learn, Transformers, LangChain, LlamaIndex, RAG, LoRA, Diffusion Models, YOLO, Recommendation Systems, NumPy, Pandas
Web Development:
React, Next.js, Angular, Node.js, Express, D3.js, React Native, HTML, CSS, JavaScript
Cloud & Tools:
AWS, GCP, Azure, Docker, MongoDB, Pinecone, Git, Flask, FastAPI, Temporal, Cassandra, DynamoDB, Spark, Airflow, AWS CDK, Terraform
Stony Brook University
- M.S. in Computer Science (Data Science & AI), GPA: 3.62
- B.S. in Computer Science, GPA: 3.71 (Magna Cum Laude)
- Email: atharvakadam@gmail.com
- LinkedIn: linkedin.com/in/atharvakadam
- GitHub: github.com/atharvakadam
