Surya Prakash Mishra

About Surya Prakash Mishra

I\'m an Applied Scientist and Data Scientist who bridges the gap between cutting-edge research and production-grade machine learning systems. With 3+ years of experience building ML solutions that drive measurable business impact, I specialize in turning complex algorithms into scalable, reliable systems. What I Do:• Design and deploy novel deep learning architectures (published researcher in computer vision and generative AI)• Build production ML pipelines using PySpark, MLFlow, and TensorFlow Extended for distributed training and real-time inference• Architect big data systems processing 500K+ events daily with Kafka, Spark Streaming, and Data Lakes• Develop statistical models and time-series forecasting solutions that drive business decisions (17% revenue increase) Recent Highlights:• At Axis Bank, built customer propensity models and pricing optimization frameworks using gradient boosting and autoregressive time-series models (ARIMA, Prophet, SARIMA) with 92% forecasting accuracy• Developed conditional diffusion models (DDPM) for synthetic microstructure generation — bridging materials science and AI• Created Vision Transformer models outperforming CNNs by 18% for defect detection in industrial applications• Designed memory-efficient multilingual tokenizers at Amazon Alexa, reducing vocabulary size by 35% while maintaining phonetic coverage across 15+ languages Technical Toolkit:ML/AI: Deep Learning, Statistical Modeling, Computer Vision, NLP, Generative AI (Diffusion Models, GANs), Transfer LearningMLOps: PySpark ML, MLFlow, TFX, Model Deployment, Experiment Tracking, A/B Testing, Distributed TrainingBig Data: Apache Spark, Kafka, Hadoop, Real-time Streaming, ETL PipelinesFrameworks: TensorFlow, PyTorch, XGBoost, Transformers (HuggingFace), OpenCVCloud: AWS (SageMaker, Lambda, S3, EC2), Docker, Kubernetes Recognition:• Ranked #22 globally in Keras contributions• Published 3 research papers in computer vision and generative AI• Axis Champions Award 2024, Samsung Hackathon Runner-Up, OpenCV AI Competition Finalist, Honda Hackathon Runner-UpI\'m passionate about solving complex problems at the intersection of research and engineering. Whether it\'s building novel architectures or scaling ML systems to production, I thrive on challenges that require both theoretical depth and practical implementation skills.Let\'s connect if you\'re working on interesting problems in ML, AI research, or data science at scale! Always open to discussing new ideas and collaborations.

Experience

1. Senior Manager

   Axis Bank

   Apr 2025 — Present · Bengaluru, IN

   Portfolio Optimization: Developed customer lifetime value (CLV) prediction models using clustering and mathematical optimization, improving portfolio segmentation and revenue forecasting

Education

• Indian Institute of Technology (Indian School of Mines), Dhanbad

  Bachelor of Technology, Mineral and Metallurgy Engineering

  2019 — 2023

• DAV Public School Chandrasekharpur

  High School, Science

  2016 — 2018

• Chinmaya vidyalaya Chennd Colony

  School

  2006 — 2016