profile-pic

Abhinav Thakur

I’m a Machine Learning and AI professional with a B.Tech in Mechanical Engineering from IIT Kanpur. I specialize in building scalable AI and deep learning solutions across domains such as computer vision, NLP, and large language models. Over the years, I’ve worked with organizations like Nanonets, Tokopedia (GoTo), EXL, and Quantiphi, where I designed and deployed end-to-end ML systems, improved product search and ranking models, and built production-ready pipelines on cloud platforms like AWS and GCP. My work has focused on solving real-world business problems through data-driven solutions, from optimizing e-commerce search systems to developing intelligent document processing and computer vision applications.

  • Role

    Senior Deep Learning Engineer

  • Years of Experience

    7.17 years

Skillsets

  • Flask
  • TensorRT
  • PyTorch
  • pandas
  • OpenCV
  • NumPy
  • NLTK
  • LLM
  • Keras
  • GCP
  • Python
  • Airflow
  • A/B testing
  • Machine Learning
  • AWS
  • Deep Learning
  • MLOps
  • Computer Vision
  • TensorFlow
  • NLP

Professional Summary

7.17Years
  • Sep, 2025 - Present 7 months

    Technology Solutions Lead II

    Gemini Solutions Pvt Ltd
  • Jan, 2025 - Aug, 2025 7 months

    Senior ML Engineer II

    Gemini Solutions Pvt Ltd
  • Sep, 2023 - Apr, 2024 7 months

    Senior Deep Learning Engineer

    Nanonets
  • Oct, 2021 - May, 2022 7 months

    Lead Assistant Manager, AI/ML

    EXL
  • May, 2022 - Sep, 20231 yr 4 months

    Senior Data Scientist

    Tokopedia
  • Jun, 2022 - Sep, 20231 yr 3 months

    Senior Data Scientist

    Tokopedia
  • May, 2019 - Aug, 20212 yr 3 months

    Machine Learning Engineer

    Quantiphi

Work History

7.17Years

Technology Solutions Lead II

Gemini Solutions Pvt Ltd
Sep, 2025 - Present 7 months

Senior ML Engineer II

Gemini Solutions Pvt Ltd
Jan, 2025 - Aug, 2025 7 months

Senior Deep Learning Engineer

Nanonets
Sep, 2023 - Apr, 2024 7 months
    Designed text extraction workflows for financial documents. Leveraged Vision and NLP techniques to reduce manual extraction time by up to 90%. Fine tuned and deployed an internal LLM model for oneshot/zeroshot text extraction.

Senior Data Scientist

Tokopedia
Jun, 2022 - Sep, 20231 yr 3 months

Senior Data Scientist

Tokopedia
May, 2022 - Sep, 20231 yr 4 months
    Led ML solution designing for search and ranking of products in e-commerce Ad-tech space. Successfully A/B tested the solution with a positive impact of $ 1 million increase in profit and 5% increase in CTR. Optimized solution for real-time latency and took it to production. Created MLOps pipeline for orchestrating and maintaining the ML solution.

Lead Assistant Manager, AI/ML

EXL
Oct, 2021 - May, 2022 7 months
    Designed and developed multiple ML based PoCs leveraging Computer Vision and NLP techniques. Built scalable ML pipelines on cloud for high volume data processing.

Machine Learning Engineer

Quantiphi
May, 2019 - Aug, 20212 yr 3 months
    End-to-end delivery from requirements gathering, solution designing, developing to deploying for clients from different industries like healthcare, finance, social media and transport. Leveraged deep learning techniques in the field of image/text analytics. Experienced in deploying ML solution architectures on cloud and embedded devices.

Major Projects

4Projects

E-commerce Search Ads

    Low latency search engine which ranks products to maximize relevance and revenue. Developed a twin tower architecture by fine tuning DistilBERT to optimize for relevance and revenue. Used a tree-based architecture to reduce infra costs and designed an MLOps pipeline for automation.

Video Analytics Solution

    Real time video analytics on an embedded system for a locomotive, including object detection and image classification models. Deployed pipeline on NVIDIA Jetson with low latency.

Intelligent Document Processing

    A scalable pipeline for text/table extraction from scanned financial documents and emails using CNN, transformers, and LLMs.

Genome Sequencing

    Latency optimized CNN model for identifying DNA sequence on cloud, reduced parameters and inference time significantly using TensorRT optimization.

Education

  • B.Tech (Mechanical Engineering)

    Indian Institute of Technology, Kanpur (2019)

  • CBSE

    Indraprastha International School, Delhi (2014)