Amit Kumar Gupta
Data Scientist
Bangalore North
Summary
Accomplished Data Scientist especially in Natural Language Processing (NLP) with a passion for delivering valuable data driven projects through analytical functions and Predictive modeling using ML. Have worked with structured and unstructured data and proficient in building NLP pipeline.
Overview
6
6
years of professional experience
7
7
years of post-secondary education
3
3
Certifications
Work History
Product Engineer-NLP
Vijna Labs Pvt. Ltd
Bangalore
02.2023 - Current
- Worked on textual documents such as structured MS-Excel sheets and unstructured text files such as
- Replaced legacy rule based system with an AI solution with performance enhancement from 35% to 90%.
- Conducted descriptive and inferential statistics, hypothesis testing, and data visualization for model validation
- Worked with and applied various deep learning and machine learning models such as RNN, LSTM, Transformers, BERT, RoBerta, XGBooost
- Experimented and built POC using Generative AI, prompt engineering on LLM such as Llama 2 for text extraction. Did fine-tuning on FLAN-T5 LLM model using PEFT/LoRA technique on AWS sagemaker.
- Collaborated with cross-functional teams such as software engineering and data engineering and ensured seamless code integration using Git.
- Tested completed projects for functionality and implemented changes to production methods to rectify issues in final products.
Data Scientist - While Doing Masters
Ensun GmbH
Siegen
10.2022 - 01.2023
- Utilized Selenium WebDriver for web scraping and textual data using MongoDB query for ML models
- Built ETL pipeline using web scrapping and MongoDB as sources
- Trained and Fine-tuned deep learning Models such as BERT and T5.
- Automated rule-based tasks
- Conducted failure analysis, identified issues, and implemented corrective measures
- Collaborated closely with software and Data Engineering teams to build and deploy pipelines
- Also conveyed complex modeling concepts and findings to both technical colleagues and non-technical stakeholders
- Adhered to Agile methodologies and tracked tasks using JIRA, while ensuring code versioning with
- GitHub CI/CD.
Research Scientist - While Doing Masters
Ciena India Pvt. Ltd
Pune
03.2022 - 10.2022
- Collaborated on the development of Action Recommendation Engine.
- Implemented prototyping (POC) rapidly and effectively for detecting alarms and anomalies in time series data, utilizing KMeans and DBSCAN clustering algorithms
- Conducted root cause analysis from failures using confusion metrics and predictive analytics using XGBoost
- Collaborated with clients and subject matter experts, presenting ideas in weekly meetings.
- Conveyed and did brainstorming on the ideas for further improvements and extracted valuable insights from ticket logs to correlate and address root causes.
Programmer Analyst
Cognizant Technology Solutions Pvt. Ltd
Bengaluru
11.2017 - 08.2019
- Actively participated in an agile environment, demonstrating exceptional adaptability by efficiently accommodating daily changes in project requirements during sprints
- Conducted extensive Exploratory Data Analysis (EDA) and applied advanced feature engineering techniques to customer data, unveiling crucial insights and patterns that informed data-driven decision-making
- Collaborated seamlessly with the team to conceptualize, develop, and implement a cutting-edge recommendation system for personalized product recommendations to users, enhancing the user experience
- Proficiently executed CRUD (Create, Read, Update, Delete) operations on customer data using SQL, ensuring data integrity and alignment with business objectives.
Education
M.Sc - Engineering
Universität Siegen
10.2019 - 04.2023
B.Tech - Mechanical Engineering
KIIT University
08.2013 - 05.2017
Skills
Python
PytorchSQLDjangoGitScikit-LearnNumpyPandasAWS ML
Generative AI with LLM
MLOps(Airflow, MLflow)
Power BI
CI/CD
Accomplishments
- Made a custom evaluation metrics for recommendation system combining Jaccard and cosine similarity to evaluate the model results
- Engineered and implemented an innovative algorithm for automatic query tag generation, revolutionizing the way relevant information is delivered
- This groundbreaking solution substantially enhanced user experience and engagement
- Exhibited a profound understanding of data enrichment techniques by developing intricate ontologies within textual data
- Employing advanced tools like Wordhoard and Conceptnet, these enrichments demonstrated an unparalleled depth of knowledge and expertise
- Stood at the forefront of technology integration by leveraging state-of-the-art NLP models, including BERT for classification tasks and harnessing the power of the GPT-3 API for text generation
Certification
Data Science by Univ.Ai
Additional Information
References
- Benjamin Hill, CEO - Ensun GmbH Siegen
https://drive.google.com/file/d/1AqYnO_0RAiyOihQsCNbTFoRN4XtZ16bk/view?pli=1
- Hasan Abu Rasheed - Research Associate - University of Siegen, Germany
https://drive.google.com/file/d/1qnwkyyBifBuzC095dTl-yX8Hd8-MydCx/view
Timeline
Generative AI with LLM - DeepLearning.Ai/Coursera
10-2023
AWS ML - Coursera
02-2023
Product Engineer-NLP
Vijna Labs Pvt. Ltd
02.2023 - Current
Data Scientist - While Doing Masters
Ensun GmbH
10.2022 - 01.2023
Research Scientist - While Doing Masters
Ciena India Pvt. Ltd
03.2022 - 10.2022
Data Science by Univ.Ai
01-2022
M.Sc - Engineering
Universität Siegen
10.2019 - 04.2023
Programmer Analyst
Cognizant Technology Solutions Pvt. Ltd
11.2017 - 08.2019
B.Tech - Mechanical Engineering
KIIT University
08.2013 - 05.2017