Amit Kumar Jha

About

Highly analytical and results-driven Data Scientist with a strong foundation in Machine Learning, Deep Learning, and Artificial Intelligence. Proven ability to develop and deploy advanced models for recommendation systems, natural language processing, computer vision, and drug discovery. Adept at leveraging quantitative analysis and feature engineering to drive significant improvements in business metrics, including sales, engagement, and operational efficiency. Seeking to apply expertise in data-driven innovation to complex challenges.

Work Experience

Advanced Data Analytics Intern

Cargill

May 2024 - Jul 2024

Contributed to data-driven solutions for agribusiness, focusing on enhancing sales and operational efficiency through advanced analytical models.

Engineered a recommendation model for farmers, ranking the top 15 farmers per sales representative using five years of data, resulting in a 12% increase in sales.
Enhanced an XGBRanker-based model through 30+ iterations, significantly improving ranking accuracy (MAP: 27%, NDCG: 30%) and increasing lead conversions by 15%.
Integrated Click-Through-Rate models (DeepFM, xDeepFM, DIN) with A/B testing, achieving a 52% NDCG score and boosting CTR by 18%.
Optimized model performance through sophisticated feature engineering techniques, reducing inference time by 25% and enhancing predictive accuracy.

Data Scientist Intern

AulaCube

Sep 2023 - Mar 2024

Developed and deployed AI-driven solutions to revolutionize school management, focusing on automation, engagement, and operational efficiency.

Spearheaded the development of AI-driven prototypes, including a Facial Recognition System, NLP-based chatbot, and a behavioral analytics tool, reducing manual administrative work by 30%.
Architected and deployed AulaGenie, a learning chatbot that processed over 500 queries daily, leading to a 25% increase in user engagement.
Optimized AulaVoz, a speech model leveraging OpenAI Whisper and Wav2Vec 2.0, improving accuracy by 28% and decreasing feedback time by 40%.

Education

Materials Science & Computer Science

Indian Institute of Technology Bombay

Nov 2021 - May 2025

Courses

Machine Learning
Deep Learning
Natural Language Processing
Computer Vision
Data Structures and Algorithms
Probability and Statistics

Projects

Boundary Smoothing for NER

Mar 2024 - Apr 2024

Developed an innovative boundary smoothing technique to enhance Named Entity Recognition (NER) model performance.

Drug Discovery Using AI

Jan 2024 - Apr 2024

Developed AI-driven models for drug classification and novel molecular structure generation, receiving a perfect 10 score for project performance.

Enhance Low-Resolution Images Using GANs

May 2023 - Jul 2023

Constructed a Generative Adversarial Network (GAN)-based super-resolution model to significantly improve image quality from low-resolution inputs.

Awards

Letter of Recommendations

Cargill & AulaCube

Jul 2024

Received letters of recommendation from both internship experiences (Cargill and AulaCube) in recognition of exemplary performance and significant contributions.

JEE Advanced 2021

Joint Entrance Examination (JEE) Board

Jan 2021

Achieved a top 2.50% ranking among over 150,000 eligible candidates in the highly competitive JEE Advanced 2021 examination.

JEE Mains 2021

Joint Entrance Examination (JEE) Board

Jan 2021

Scored 99.37 percentile among over 940,000 eligible candidates in the national-level JEE Mains 2021 examination.

Skills

Programming Languages

Python
C++

Machine Learning Frameworks

PyTorch
TensorFlow
Scikit-Learn
XGBoost

AI & Machine Learning

Deep Learning
Transformers
Natural Language Processing (NLP)
Computer Vision
Recommendation Systems
Generative AI
Neural Networks

Data Tools & Databases

SQL
OpenCV

MLOps & Cloud

MLOps
Docker
CI/CD
AWS

Research & Optimization