0% found this document useful (0 votes)
71 views1 page

20HS20067

Suraj Gupta is an M.Sc. student in Economics at IIT Kharagpur with a minor in Mathematics and Computing, and a specialization in Artificial Intelligence. He has completed multiple internships focusing on AI/ML and data analytics, achieving significant improvements in model performance and fraud detection. Suraj has also founded a technology society, held leadership positions, and received academic awards for his achievements.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
71 views1 page

20HS20067

Suraj Gupta is an M.Sc. student in Economics at IIT Kharagpur with a minor in Mathematics and Computing, and a specialization in Artificial Intelligence. He has completed multiple internships focusing on AI/ML and data analytics, achieving significant improvements in model performance and fraud detection. Suraj has also founded a technology society, held leadership positions, and received academic awards for his achievements.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

SURAJ GUPTA | 20HS20067

ECONOMICS ([Link]. 5Y)


MINOR in MATHEMATICS & COMPUTING ([Link]. 5Y)
MICRO SPL. in ARTIFICIAL INTELLIGENCE AND APPLICATIONS
EDUCATION
Year Degree/Exam Institute CGPA/Marks
2025 [Link](5YR) IIT Kharagpur 8.32 / 10
2020 Class XII KES Junior College of Science, Mumbai 85.38%
2018 Class X Abhyudaya Nagar School, Mumbai 92.6%
INTERNSHIPS
AI/ML Intern | Elucidata Data Consulting Pvt. Ltd. | New Delhi Jan '24 - Jun '24
Extracted biomedical entities - diseases, drugs, genes and their relationships, stored in a Graph DB using RAG techniques for query retrieval
• Utilized Ollama Embeddings and ChromaDB for query retrieval using open-source models like Llama3, Gemma2 and Mistral via Groq API and Ollama
• Created a Knowledge Graph-based retrieval system using Neo4j Graph Database and LangChain by normalizing entities from 3000+ publication papers
• Developed autonomous AI agents using AutoGen and function calling to generate accurate cypher queries, integrating the output to a chainlit chatbot
• Implemented advanced RAG system using GraphRAG with gpt-4o and gemini-1.5 flash, generating over 1 lakh+ nodes & edges, reducing costs by 66%
Data Analytics Intern | EXL Services | New Delhi Jul '24 - Aug '24
Enhanced fraud detection capabilities for UK banking data, focusing on high accuracy and efficient data handling using ML and DL algorithms
• Performed extensive feature engineering with One Hot Encoding on a dataset of 5 lakh+ entries, enhancing the model performance and interpretability
• Applied SMOTE sampling technique and Robust Scaler for normalization, used z-test for feature selection, ensuring robust and reliable model inputs
• Developed the model with XGBoost & ANN, achieving an F1 Score of 0.86, an AUC of 0.92 and improving fraudulent transaction identification by 25%
Data Science Intern | Raymond Limited (Analytics Department) | Mumbai Jul '22 - Aug '22
Designed an end to end machine learning pipeline aimed at enhancing the cross-sell opportunities for trousers to shirt buyers on realtime data
• Engineered 27- dimensional feature vectors on over 1 lakh+ records, improving F2 scores from 0.72 to 0.84, utilizing XGBoost and RandomForest model
• Employed the Cross Validation technique - KFoldCrossValidation and hyperparameter tuning using RandomizedSearchCV to fine-tune the models
• Successfully deployed the model to analytics department, resulting in a increment of 3% response rate from a test on 10,000 customers within 1 week
Research Intern | Department of Industrial and Systems Engineering | IIT Kharagpur Jun '22 - Present
Conducted in-depth analysis on migration rates in Sundarbans, examining impact of economic indicators under the supervision of Prof. Sayak
• Employed Selenium to meticulously extract and merge 630+ features, from the CMIE Database using panel data regression analysis in Stata Software
• Utilized advanced econometric techniques including Hausman, LM test to determine the significant features, resulting in a refined Fixed Effects model
• Collaborating on an extensive research paper, extending the scope of the work to state-level data analysis, with 67.8 lakh datapoints and 772 features

PROJECTS
Research Project | Prof. Dana Anderson | University of Alberta Jun '23 - Jul '23
• Built a random-effects GLS model on 84 observations over 21 years, identifying key metrics significantly affecting crop yields (R² = 35.81%) in Ukraine
• Addressed multi-collinearity in models using the Hausman test, confirming the random-effects approach for optimizing crop strategies across 4 regions
Data Analytics (Bronze) | General Championship 2023 Mar '23 - Apr '23
• Developed a CNN-GRU time series forecasting model for predicting manufacturing unit states based on 25 sensors, achieving an overall 97% accuracy
• Improved inventory management by accurately predicting unit states at 10-minute intervals, achieving 1st position in accuracy and an overall podium

ENTREPRENEURIAL EXPERIENCES
Founder & Executive Head | KodeinKGP : Technology Web3.0 Society | IIT Kharagpur Apr '21 - Present
• Founded and led the KodeinKGP : Technology Web3.0 Society, developing a problem-solving web portal utilizing ReactJS for student-related challenges
• Built a talented team of executives in AI, Metaverse, Blockchain and Events facets, raising INR 5 lakh+ amount through sponsorship from 8 companies
• Established and managed a thriving blockchain community with 2500+ participants, consistently sharing weekly blogs and posting relevant opportunities

AWARDS AND ACHIEVEMENTS


• Awarded the Technology Alumni Association KGP Chapter Scholarship, recognized as the academically best student of 3rd year among 1800+ students
• Secured a department change in first year being among the top 5% out of 1800+ students to switch to the Economics Branch of the Department of HSS

SKILLS AND EXPERTISE


Industry Domain Knowledge: Data Visualization, Data Analytics, Machine Learning, NLP, Deep Learning, Neural Nets, Generative AI, Artificial Intelligence
Languages & Frameworks: Python, C++, Llama3, Gemini, OpenAI, AutoGen, Chainlit, Docker, LSTM, GRU, RNN, CNN, Transformer, Langchain, SQL, MySQL
Tools & Technologies: Git, Github, Numpy, ScikitLearn, Tensorflow, Keras, Linux, Nvidia NIM, Ollama, Huggingface, AWS, Neo4j AuraDB, Multimodal RAG
Relevant Courses: Econometric Analysis, File Organization and Database Systems, Probability and Statistics, Programming and Data Structures, Artificial
Intelligence, Operation Research, Corporate Valuation, Financial Institutions and Markets, Machine Learning, AI/ML, Mircroeconomics, Macroeconomics

POSITIONS OF RESPONSIBILITY
Hall President | Vidysagar Hall of Residence | IIT Kharagpur Jul '23 - Apr '24
• Chaired a 3-tier student council of 9 General Secretaries and 19 Secretaries; co-ordinated with the warden, managing 18 lakh+ budget (20% YoY increase)
• Member of Students' Senate, the apex student body at IIT Kharagpur, ideating and executing key decisions catering to 0.17 lakh+ students of the institute
Secretary : Web | Technology Students' Gymkhana | IIT Kharagpur Jul '21 - Jul '22
• Built a Digital Inventory System to manage and record the transactions of the sports inventory of around 6 stores in IIT Kharagpur worth INR 10 crores
• Spearheaded the development of the official Gymkhana website, incorporating certificate generator and notifications features from the ground up

EXTRA CURRICULAR ACTIVITIES


Sports and Games: Represented Vidyasagar Hall of Residence in the table-tennis team in Interhall Sports Inter Hall Competition 2021-2022, IIT Kharagpur
Adventure: Organized and Participated in Har Ki DunTrek, Uttarakhand and Sar Pass Trek, Himachal Pradesh as Governor, Technology Adventure Society
!Self declared by the student, CDC could not verify the relevant documents

You might also like