Nishant Kadian
Brampton
Summary
Data Scientist with MS degree in Data Science and 4 years of experience in Python, Machine learning , Large Language Models( LLAMA,ChatGPT etc), NLP, Deep Learning, Computer Vision (Face Recognition Models), model deployment on cloud platforms like AWS, Databses- SQL, MongoDB,ChromaDB. Experienced in leading and mentoring junior data scientists, and collaborating with stakeholders and cross-functional teams to identify opportunities and drive data-driven strategies.
Overview
6
6
years of professional experience
Work History
Senior Data Scientist
HT Media Ltd
2023.11 - 2024.03
- Handled projects at HT Media's flagship job portal, Shine.com, focusing on enhancing user experience and optimizing recruitment processes.
- Developed and implemented a cutting-edge recommendation engine leveraging advanced machine learning techniques. Prepared a custom dataset and fine-tuned the LLAMA2-7B large language model on this dataset using PEFT techniques like LORA. Implemented XGBoost classifier on top of the model's embeddings to create precise candidate recommendations tailored to recruiters' needs.
- Created an innovative algorithm to deliver personalized job recommendations to candidates. Utilized the Falcon-7B Large Language Model, fine-tuned it on proprietary data, to generate embeddings for both job listings and candidate profiles. Engineered a seamless integration with the Chroma vector database, leveraging cosine similarity for efficient and accurate job matching. This algorithm enhanced the job search experience, enhancing candidate engagement and satisfaction.
- Established a robust data pipeline for lead generation, streamlining the process from data acquisition to lead prioritization. Integrated diverse data sources including SQL and NoSQL databases such as Solr and MySQL. Implemented preprocessing techniques and applied tailored filters to optimize lead quality. Developed priority scoring mechanisms to identify high-value leads, facilitating targeted outreach strategies. Created seamless data transfer to AWS S3, ensuring accessibility for the sales team and maximizing lead conversion potential.
Deputy Manager-Data Science
WNS Global Services
2022.04 - 2023.11
- Contributed to UK based insurance company by developing a highly accurate GLM (Generalized Linear Model) to predict loss-cost for claims.
- Provided underwriters with essential rating factors by exporting the model's relativities. These factors were Utilized in an Excel-based rating tool to calculate premiums effectively.
- Performed Sentiment Analysis (NLP) on customer's reviews on claim settlement satisfaction using cutting edge BERT transformer model in PyTorch and K-Means clustering Algorithm.
- Used OpenAl LLM by langchain library to convert some reviews from Spanish language to English language as preprocessing step before sentiment analysis.
- Guided and lead a team of junior data scientists.
Data Scientist
Mahindra Teqo
2021.09 - 2022.04
- Developed a Machine Learning Model to accurately predict and mitigate clipping loss (energy loss) in solar power plant inverters.
- Utilized unsupervised Machine Learning algorithms, specifically clustering algorithms, to identify inverters experiencing clipping loss. Implemented a supervised ML algorithm, XG Boost, to predict the exact amount of clipping loss.
- The deployed model on cloud platform AWS successfully increased the power output of the client's solar plant by an impressive 1.78%.
Data Science Consultant
Ernst Young LLP
2021.02 - 2021.09
- Developed a Machine Learning Classification Model (XG Boost Algorithm) to predict the probability of policy non-persistence at the time of application, resulting in a 1% increment in client's renewal income.
- Created and deployed an AWS-based API for seamless access to the model.
- Collaborated on the "Due Risk Scoring" project, developing an innovative ML model (XG Boost and Deep Learning model ) to predict the probability that a policy will go lapse (the customer will not pay premium within 180 days of the due date).
Data Scientist
Flip Robo Technologies
2020.06 - 2021.02
- Designed and implemented a robust face recognition system for efficient attendance monitoring of remote employees.
- Leveraged the power of FaceNet deep learning neural network to extract embeddings (vector representations) from face images.
- Utilized Cosine Similarity to compare embeddings and accurately identify individuals.
FIN & ADMIN BUSINESS ASSOCIATE
IBM
2018.05 - 2019.01
Education
Master of Science - Data Science
Liverpool John Moores University
2021-06
Bachelor of Technology - ComputerScience &Engineering
Guru Gobind Singh Indraprastha University
2017-05
Skills
- Machine learning Algorithms
- Large Language Models- LLAMA, Falcon,Open AI GPT, BERT
- Programming Skills-Python: NumPy, Pandas, matplotlib, scikit-learn
- Databases- SQL, MongoDB,Solr
- Vector Databses- ChromaDB
- Natural Language Processing
Timeline
Senior Data Scientist
HT Media Ltd
2023.11 - 2024.03
Deputy Manager-Data Science
WNS Global Services
2022.04 - 2023.11
Data Scientist
Mahindra Teqo
2021.09 - 2022.04
Data Science Consultant
Ernst Young LLP
2021.02 - 2021.09
Data Scientist
Flip Robo Technologies
2020.06 - 2021.02
FIN & ADMIN BUSINESS ASSOCIATE
IBM
2018.05 - 2019.01
Master of Science - Data Science
Liverpool John Moores University
Bachelor of Technology - ComputerScience &Engineering
Guru Gobind Singh Indraprastha University
Similar Profiles
DIVYA MALIKDIVYA MALIK
Recruiter at HT Media LtdRecruiter at HT Media Ltd
Urvesh DhangarUrvesh Dhangar
Solution Engineer at HT Media LabsSolution Engineer at HT Media Labs
Nishant DuttaNishant Dutta
Head – Alternate Channels at HT Media Ventures Ltd – OTT PlayHead – Alternate Channels at HT Media Ventures Ltd – OTT Play