Abhiram Deshmukh

Results-driven AI/ML Engineer with 4+ years of experience designing, developing, and deploying scalable machine learning, NLP, deep learning, and Generative AI solutions across enterprise environments. Skilled in Python, SQL, TensorFlow, PyTorch, Scikit-learn, XGBoost, LangChain, LlamaIndex, Hugging Face, RAG pipelines, FastAPI, Docker, Kubernetes, MLflow, CI/CD, AWS, Azure, and GCP. Experienced in building production AI/ML systems, data pipelines, model evaluation workflows, LLM applications, and stakeholder-facing analytics solutions.

AI / Machine Learning Engineer - American Express - Phoenix, AZ

(2025-08)

• Developed scalable AI/ML and deep learning models using Python, TensorFlow, PyTorch, Scikit-learn, XGBoost, and LightGBM for fraud detection, customer analytics, and predictive modeling, improving fraud detection accuracy by 28% and reducing false positives by 22%.
• Built and deployed Generative AI and LLM applications using GPT-4o, Claude, Gemini, Hugging Face, LangChain, LlamaIndex, and LangGraph for chatbot automation, document processing, and customer support workflows.
• Designed RAG pipelines using FAISS, Pinecone, Weaviate, Qdrant, and ChromaDB to enable semantic search and intelligent knowledge retrieval, increasing retrieval accuracy by 35% and reducing lookup time by 60%.
• Implemented MLOps and LLMOps pipelines using MLflow, Kubeflow, GitHub Actions, CI/CD, and LangSmith for automated training, deployment, monitoring, model versioning, and release reliability.
• Developed ETL and data pipelines using Apache Spark, Apache Airflow, SQL, PostgreSQL, MongoDB, Pandas, and NumPy to process large-scale structured and unstructured data while reducing processing latency by 45%.

Senior Data Scientist - Commonwealth Bank - Bangalore, India

(2022-04 - 2023-06)

• Developed advanced data science and ML models using Python, Scikit-learn, XGBoost, Random Forest, Pandas, and NumPy for customer analytics, risk assessment, and underwriting optimization, improving prediction accuracy by 26%.
• Designed and optimized ETL pipelines using Python, SQL, MySQL, AWS S3, and Apache Airflow to integrate, clean, and process banking datasets, reducing data preparation time by 50% and improving data quality by 30%.
• Performed EDA, statistical analysis, feature engineering, data visualization, and KPI reporting using SciPy, Matplotlib, Tableau, and Power BI to uncover insights and support stakeholder decision-making.
• Developed and deployed scalable Flask APIs, Docker-based applications, and cloud-integrated ML solutions for real-time model serving and reporting, reducing inference latency by 40% and supporting 100K+ monthly prediction requests.

Data Scientist - Toyota Kirloskar Motor - Bangalore, India

(2020-12 - 2022-03)

• Developed and deployed machine learning models using Python, Scikit-learn, and TensorFlow to analyze business data and improve predictive accuracy, increasing forecasting accuracy by 25%.
• Performed data cleaning, transformation, and EDA on large datasets using Pandas, NumPy, and SQL, improving data quality by 35% and accelerating analytics delivery by 40%.
• Built interactive Power BI and Tableau dashboards to support business stakeholders in data-driven decision-making, reducing reporting turnaround time by 60%.
• Designed and optimized SQL queries, stored procedures, and ETL pipelines for efficient extraction, transformation, and loading of structured and unstructured data, improving query performance by 50%.

Master of Science - Computer Science - University of the Pacific (2023-08 - 2025-05)

Bachelor of Technology - Electronics & Telecommunication - Vishwakarma Institute of Technology (2016-08 - 2020-05)