Nikhil Pandey
Data Scientist
Bangalore , India
Summary
Data Scientist familiar with gathering, cleaning and organizing data for use by technical and non-technical personnel. Advanced understanding of statistical, algebraic and other analytical techniques. Highly organized, motivated and diligent with significant background in computer vision, NLP, LLMs, Predictive Modeling, and Recommender Systems.
Overview
6
6
years of professional experience
4
4
years of post-secondary education
Work History
Senior Data Scientist
Protosight
Bangalore
01.2023 - Current
- Utilized advanced querying, visualization and analytics tools to analyze and process complex data sets. Stressed on laying strong data science fundamentals.
- Discovered mistakes in existing codebase and rectified them leading to increase in 9.5% increase in customer retention.
- Built company A/B testing frameworks to test model quality further leading to removal of features leading to massive 22% jump in main KPI.
- Identified, measured and recommended improvement strategies for KPIs across application space.
Lead Data Scientist
XA Group
Dubai
11.2021 - 03.2023
- System Requirements and ML Engine Design: Gathered system requirements and designed an efficient ML engine and data pipeline based on stakeholder inputs.
- Computer Vision Model Development: Created end-to-end computer vision model training, testing, and deployment pipelines from scratch. Developed models for segmentation, classification, and detection tasks.
- Performance Enhancement: Achieved an impressive 58 mAP@IoU=[.5,.95] with 25k images, improving system performance from 11 mAP. Utilized data-driven approaches to optimize models and enhance overall system efficiency.
- KPI Formulation and Evaluation Framework: Formulated Key Performance Indicators (KPIs) to monitor AI solution performance post-deployment. Established robust evaluation frameworks for ongoing performance assessment.
- Production-Grade API Development: Created high-performance APIs to return results within 0.5 seconds on the
existing platform. Developed online and offline API for image validation to ensure data integrity. - Cross-Functional Collaboration: Worked effectively in a cross-functional environment to ensure successful release of MVP 1.0. Collaborated with diverse teams to deliver impactful AI solutions.
- Team Building and Leadership: Conducted interviews and played a key role in hiring, assembling and developing the data science team.
CTO
Agrograde
Mumbai
01.2019 - 10.2021
- Team Leadership and Product Development: Managed company-level architecture, platform and data configuration processes and implementation protocols.
- Optical Sorting and Remote-QC Software: - Spearheaded engineering team in developing AI-based optical sorting machines and Remote-QC software for online commodity traders. Achieved successful implementation of product from ideation to market launch.
- Robotic Integration and High-Speed Inference: Designed an innovative architecture enabling robotic integration of AI models, delivering an impressive 35 FPS inference speed with over 95 percent accuracy with quantized models.
- Subjectivity Removal App: Conceptualized and executed creation of Agrograde’s app, utilizing computer vision and statistics to remove subjectivity, leading to enhanced value realization for users.
- Patent Authorship and State-of-the-Art Contributions: Authored and published patents for state-of-the-art optical sorting technologies, highlighting expertise and contributions to the field.
Data Scientist
Occipital Technologies
Mumbai
04.2018 - 01.2019
- Applied statistical and algebraic techniques to interpret key points from gathered data.
- Assessed accuracy and effectiveness of new and existing data sources and data analysis techniques.
- Delivered data based products which directly increased start-up's revenue by 60%.
Data Science Intern
IIT-Khargpur
Khargpur
03.2017 - 12.2017
- Developed and coded software programs, algorithms and automated processes to cleanse and evaluate large datasets from multiple disparate sources.
- Applied appropriate data science techniques to solve research problems.
- Learned nitty-gritty of Data Science and Machine Learning from top talents of country.
Education
Bachelor of Science - Computer Science And Engineering
IT-GGV
Chhatisgarh 08.2014 - 04.2018
Skills
Intelligence gathering
Machine learning
Agile framework understanding
Deep Learning
Visualization
SQL
Python
Computer Vision
NLP
Recommender Systems
Pandas
Scikit-Learn
PyTorch
Numpy
OpenCV
SpaCy
Patents
Multigrade sorting system for fruits and vegetables:
- Nikhil Pandey, Gaurav Pardeshi, Kshitij Thakur and Rakesh Barai
System for assaying quality of agricultural produce and a method thereof.
- Nikhil Pandey, Gaurav Pardeshi, Kshitij Thakur and Rakesh Barai
Publication
- Grading and Defect Detection in Potatoes Using Deep LearningGrading and Defect Detection in Potatoes Using Deep LearningCommunication, Networks and Computing. CNC 2018. Communications in Computer and Information Science, vol 839. Springer, Singapore · Mar 23, 2018Communication, Networks and Computing. CNC 2018. Communications in Computer and Information Science, vol 839. Springer, Singapore · Mar 23, 2018
- Various Techniques Used In Defect Segmentation and Disease Inspection in Fruits and Vegetables: A SurveyVarious Techniques Used In Defect Segmentation and Disease Inspection in Fruits and Vegetables: A Survey Volume 5, Issue X, International Journal for Research in Applied Science and Engineering Technolog · Oct 1, 2017
Timeline
Senior Data Scientist
Protosight
01.2023 - Current
Lead Data Scientist
XA Group
11.2021 - 03.2023
CTO
Agrograde
01.2019 - 10.2021
Data Scientist
Occipital Technologies
04.2018 - 01.2019
Data Science Intern
IIT-Khargpur
03.2017 - 12.2017
Bachelor of Science - Computer Science And Engineering
IT-GGV
08.2014 - 04.2018