Anurag M.
GenAI Engineer/ AI Engineer/ Data Scientist
Notable Highlight
Built memory-driven agentic LLM systems and RAG pipelines, achieving 30–50% token savings and 90.1% accuracy. Optimized Amazon adTech campaigns with ML pipelines that improved ROAS by 42%.
Experience Summary
AI Engineer with hands-on expertise in agentic LLM systems, RAG pipelines, and ML optimization. Joined Sig. as Associate Data Scientist, building adTech ML pipelines. Previously interned at TA on neural network research. IIT Bhubaneswar dual degree graduate with strong foundation in transformers, PyTorch, and production AI systems.
Key Achievements
- Built machine learning pipelines and data optimization solutions for enterprise client environments. - Developed expertise in Applied AI technologies, including NLP, Retrieval-Augmented Generation (RAG), Agentic AI systems, and MCP server implementations. - Contributed to advertising technology initiatives focused on revenue optimization through machine learning and data-driven strategies. - Gained end-to-end exposure to e-commerce customer journeys through Amazon DSP-related projects, analyzing user behavior from ad engagement through conversion. - Worked with healthcare and pharmaceutical clients to support data science and analytics initiatives. - Developed AI-powered chatbot and conversational AI solutions designed to improve information retrieval and user engagement. - Transitioned from traditional data science into Generative AI and Applied AI specializations. - Graduated from the Indian Institute of Technology (IIT BHU). - Demonstrated adaptability by continuously upskilling in evolving AI technologies and frameworks.
Skills
Work Experience
Associate Data Scientist
Sig. · Jul 2025 – Present
- Built ML-driven optimization pipelines for Amazon adTech campaigns, improving ROAS by 42%, impressions by 55%, and DPVR by 36% across client campaigns.
- Developed scalable Python pipelines and reusable API-integrated automation workflows for large-scale advertising datasets, reducing manual reporting effort and improving operational efficiency.
Data Science Research Intern
TA · May 2024 – Jul 2024
- Developed PyTorch-based experimentation pipelines, running 10+ model configurations for performance optimization.
- Worked on Binary Neural Networks (BNNs), studying scalability and efficiency of larger BNN architectures (3B+ parameters) compared to standard neural networks.