Personal details

VINOD M. - Remote data scientist

VINOD M.

Based in: 🇬🇧 United Kingdom
Timezone: London (UTC+1)

Summary

An experienced data scientist with extensive technical proficiency and a consulting background gained from diverse projects across medical, retail, consumer, space, and energy domains. My focus lies in applying data science to tackle complex challenges and providing innovative solutions that yield significant business outcomes. I possess a distinctive combination of technical prowess and strategic thinking, underpinned by a robust foundation in fields like statistical analysis, machine learning, deep learning, NLP etc. Currently, I am actively pursuing challenging opportunities to further enhance my technical skills, aiming to contribute to a forward-thinking organization where my expertise can propel development and technological advancement.

Work Experience

Data Scientist
Sagentia Innovation | Sep 2022 - Present
Computer Vision
Data Science
NLP (Natural Language Processing)
Statistical Analysis

• Collaborated with clients to identify critical business challenges and implemented data-driven solutions, utilizing advanced AI methods for automated problem-solving and increased operational efficiency. • Designed a machine learning model to predict disease onset based on sleep patterns and deliver personalized sleep recommendations through long-term data analysis. • Coded and designed a predictive model that predicts the amount of carbon release in the future from a manufacturing unit which helped in altering the company's net-zero strategy using detailed research on the UK energy market. • Led an advanced data analytics project exploring the impact of biomarkers on aging and intestinal health using machine learning and statistical techniques. A 35% improvement in strategic planning was made possible. • Used strong presentation and effective communication skills to explain complex analytical findings to stakeholders and non technical teams.

Data Scientist
Pass_by | Jan 2021 - Aug 2022
Google BigQuery
Google Cloud Platform
Deep Learning
Airflow

Spearheaded the development of end-to-end data pipelines to deliver results in dynamic and collaborative environments by productionising the solution with specific code design. • Engaged in the extracting, cleaning, manipulating, and preprocessing of large datasets from multiple sources to identify patterns and trends using Python programming. • • Managed a data science pipeline for forecasting and assessing customer behavior, resulting in a 30% boost in sales and facilitating efficient product development. • Derived actionable insights from complex datasets which contributed to meaningful improvements to the existing models through directed research work and business decision-making. • Partnered with functional teams to understand business needs, and delivered analytical reports tailored to targeted audiences.

Education

University of Bath
Master's degree・Data Science
Sep 2020 - Sep 2021
Visvesvaraya Institute of Technology
Bachelor's degree・Electronics and Communication
Aug 2013 - Jul 2017

Personal Projects

Beacon object detection
2022
Image Processing
Azure
Computer Vision
This project was led by me to develop a proof of concept for live object detection on farming machinery using modern computer vision technology. The YOLOV5 model is used for object detection of classification of different beacons with respect to their position. Deployed the model as a web application using Flask and Microsoft Azure to demonstrate a live demo of how it will work when it's deployed on the live system. Collaborated with a cross-functional team at the time of implementation for collecting the data and was directly involved with clients to understand business requirements.
NLP patent classification tool
2022
Python
NLP (Natural Language Processing)
Text Mining
The project's main objective was to use Natural Language Processing (NLP) techniques to speed up and improve the patent classification process. The pretrained BERT model has been used and fine-tuned on the labeled patent dataset. During fine-tuning, the model learns to map the input patent text to the corresponding categories. The model's performance has been evaluated using performance metrics such as accuracy, precision, recall, and F1 score to tell how effectively it classifies patents. Overall, the project reduced manual classification time by 50% and increased team productivity by 30%.