Saumit Paul
Ahmedabad
Overview
4
4
years of professional experience
Work History
Data Scientist
DeepTek.ai
Pune
2022.06 - Current
- Developed Computer Vision models on Chest X-Rays for detecting and segmenting various abnormalities
- Improved Pleural Effusion model's AUROC score from 0.82 to 0.91 on out-of-sample test data with significant improvement in classification metrics.
- Worked on 2 successful PoC's for a potential offshore client. The projects included classification for histopathological scans and multi-label classification of various types of arrhythmia in dog ECG signals.
- Conducted statistical AI validation studies for regulatory approval applications from US FDA (cleared), Thai FDA (cleared) , HSA , CE etc.
- Responsible for internships within the data science team. This includes hiring, mentoring and working closely with interns.
Analytics & Automation Intern
Anheuser-Busch InBev
Bengaluru
2022.01 - 2022.05
- Developed Power BI Dashboard to visualize change in working patterns of employees in hybrid and remote work settings.
- Built and maintained Data Pipeline and SQL Database on Azure for the dashboard.
- Performed data wrangling on multiple data points fed from the Task Mining API.
Data Science Intern
iQGateway
Bengaluru
2021.07 - 2021.08
- Worked on a model diagnostic visualization feature for a project aimed at providing a service to facilitate AutoML Pipelines.
Machine Learning Intern
Prime Focus Technologies
Bengaluru
2020.08 - 2020.10
- Wrote Python Script to facilitate the end-to-end process from mixed audio input to separate audio output
- Tested open-source Deep Learning model performance on the task of Audio Source Separation on custom audio data
- Learnt about Audio Signal Processing methods for feature extraction & their implementation in Python.
Education
B.Tech - Electrical & Electronics Engineering
Manipal Institute of Technology
Manipal, Karnataka07.2022
Skills
- Data Analysis / Statistics in Python (NumPy, Pandas, SciPy, Matplotlib, Seaborn, OpenCV, Plotly, Streamlit)
- Deep Learning / Machine Learning in Python (TensorFlow, Keras, Scikit-Learn, MLFlow)
- MongoDB
- MySQL
- Git
Publications
- A Comprehensive Evaluation of DeepTek CXR Analyzer in Detecting and Localising Suspicious Findings in Chest X-rays (https://dx.doi.org/10.26044/ecr2024/C-23322)
Timeline
Data Scientist
DeepTek.ai
2022.06 - Current
Analytics & Automation Intern
Anheuser-Busch InBev
2022.01 - 2022.05
Data Science Intern
iQGateway
2021.07 - 2021.08
Machine Learning Intern
Prime Focus Technologies
2020.08 - 2020.10
B.Tech - Electrical & Electronics Engineering
Manipal Institute of Technology