Headline

Top ML Voice | Datascience @EY |Machine Learning | GenAI | Agentic AI

About

AI/ML Engineer and Consultant at Ernst & Young. ✨ I enjoy working on hard problems 🧩: the kind that sit between research and production, where the architecture decisions actually matter.👩🏻‍💻 Over the last four years, I've worked on: • Multi-agent AI systems and LLM orchestration at enterprise scale • Multi-modal RAG pipelines- BFSI NBFC Domain • Supervised and unsupervised ML — customer segmentation, classification, cross-sell recommendation • NLP systems for government infrastructure — large-scale ticket classification, intent detection, and real-time complaint consolidation • End-to-end MLOps — automated retraining, SageMaker inference, drift detection, and A/B testing 🍂National finalist, Aadhaar AI Hackathon. Python Gold on HackerRank. Presented research at the 46th International Systems Conference. TensorFlow Developer certified.

Location

India

AI stack

cursor

opencode

claudecode

Work experience

Consultant at EY, Sep 2025 - Present · 1 yr

Associate Consultant at EY, Oct 2024 - Sep 2025 · 1 yr

Data Science and ML at EY, Jun 2023 - Oct 2024 · 1 yr 5 mos, Built cross-sell recommendation engine using a hybrid of supervised and unsupervised (clustering) algorithms for customer segmentation, with end-to-end feature engineering pipeline across 15+ product lines. Deployed full MLOps pipeline on Amazon SageMaker — automated retraining, model drift detection, SageMaker inference endpoints, and A/B testing framework; improved cross-sell conversion rate by ~22%.

Data Science at EY, Feb 2023 - Apr 2023 · 3 mos, New Delhi, Delhi, India, Worked on Government Projects associated with Skill Development. Identified, analyzed, and interpreted trends in complex data sets by finding correlations and visualizing charts. Worked on Pradhan Mantri Kaushal Vikas Yojana Tickets Dataset. Performed Topic modeling and other NLP based Textual Analysis techniques such as Tokenization and N gram modeling

Machine Learning intern at Genisup India Pvt. ltd., Apr 2022 - Aug 2022 · 5 mos, Bangalore Urban, Karnataka, India, Advanced Image Classification Implemented EfficientNet B7 architecture on the Caltech CUB-200 dataset, expanding its size and achieving state-of-the-art performance. Demonstrated expertise in fine-grained image classification and advanced CNN models, including RESNET50, VGG16, and Inception V3. Utilized transfer learning techniques to leverage pre-trained models and improve accuracy. Applied image processing techniques such as 2D convolution and low pass filters for image quality enhancement. Machine Learning-Based Classification Implemented various classification algorithms including Decision Trees, Random Forest, XGBoost, and ADABoost to achieve accurate predictions. Improved model performance by utilizing techniques like Cost Complexity Pruning and OOB error analysis to optimize accuracy and minimize errors.

Artificial Intelligence and Machine Learning at 7Classes, Jul 2020 - Sep 2020 · 3 mos, Mumbai, Maharashtra, India, Applied and Learned Data Analysis, Gitlab, DBMS, Various Machine Learning Algorithms, basics of NLP, Also I have created one model to recommend test on site, based on student's previous attempts, this code was successfully deployed on 7classes. All of my work was done under the guidance of trainer and C.E.O. of 7 classes and instaprep Mr. Anup Raaj.

Education

B.tech, Electrical and Computer Science, Dayalbagh Educational Institute, Agra

Highschool and Intermediate, I.P.S. Nagaur Rajasthan

Public links

github: https://github.com/sumiranbhatnagar