Anwesha Samaddar — Data Scientist

I am a data science and analytics professional with over three years of experience spanning research, industry, and academia. I previously worked as a Research Assistant at the Center for Magnetic Resonance Research (UMN, Twin Cities), where I applied deep learning models for medical image segmentation and motion correction in MRI scans. Additionally, I served as a Quantitative Analyst at Credit Suisse, focusing on time-series forecasting, financial risk model validation, and predictive analytics.I have also completed multiple internships, including my role at American Family Insurance, where I optimized computer vision models for odometer extraction, and at a Fintech startup where I developed OCR and deep learning solutions for document verification. Alongside my professional roles, I am a Teaching Assistant at the Carlson School of Management, mentoring over 100 students in data visualization, A/B testing, regression, time-series forecasting, and optimization using R, Power BI, and Excel.With expertise in Python, R, SQL, machine learning, and data visualization, I bring a strong analytical mindset and a passion for leveraging data-driven insights in finance, healthcare, and technology applications.

Stackforce AI infers this person is a Data Science and Analytics professional with strong expertise in Fintech and Healthcare.

Location: Madison, Wisconsin, United States

Experience: 4 yrs 5 mos

Skills

Large Language Models (llm)
Agentic Ai Development
Convolutional Neural Networks (cnn)
Applied Machine Learning
Data Visualization
Statistical Data Analysis
Image Processing
Diffusion Tensor Imaging
Optical Character Recognition (ocr)
Computer Vision
Quantitative Analytics
Data Analytics
Analytical Skills
Machine Learning Algorithms
Programming Languages

Career Highlights

Over three years of experience in data science and analytics.
Expertise in deep learning for medical image processing.
Strong analytical mindset with a focus on data-driven insights.

Work Experience

American Family Insurance

Data Scientist (9 mos)

University of Minnesota School of Dentistry

GenAI Developer (2 mos)

American Family Insurance

Machine Learning Intern (3 mos)

University of Minnesota - Carlson School of Management

Graduate Teaching Assistant (1 yr 5 mos)

University of Minnesota Center for Magnetic Resonance Research

Graduate Research Assistant (11 mos)

University of Minnesota College of Science and Engineering

Graduate Student - AI and Robotics (1 yr 11 mos)

Credit Suisse

Quantitative Analyst (1 yr 1 mo)

Yatiken Software Technologies

NLP Intern (1 mo)

Smiles.ai

Computer Vision (2 mos)

Salary Dost

OCR and Deep Learning (5 mos)

Department of Mechanical Engineering, IIT Kharagpu

Secretary at Mekanika (8 mos)

Education

Master of Science - MS at University of Minnesota College of Science and Engineering

Master of Technology - MTech at Indian Institute of Technology, Kharagpur

Anwesha Samaddar

Data Scientist

Madison, Wisconsin, United States4 yrs 5 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Over three years of experience in data science and analytics.
Expertise in deep learning for medical image processing.
Strong analytical mindset with a focus on data-driven insights.

Stackforce AI infers this person is a Data Science and Analytics professional with strong expertise in Fintech and Healthcare.

Contact

Skills

Core Skills

Large Language Models (llm)Agentic Ai DevelopmentConvolutional Neural Networks (cnn)Applied Machine LearningData VisualizationStatistical Data AnalysisImage ProcessingDiffusion Tensor ImagingOptical Character Recognition (ocr)Computer VisionQuantitative AnalyticsData AnalyticsAnalytical SkillsMachine Learning AlgorithmsProgramming Languages

Other Skills

Retrieval-Augmented Generation (RAG)Python (Programming Language)Neuro-Linguistic Programming (NLP)GaNOpenAIHuggingfaceMicrosoft ExcelRjamoviPyTorchC++Generative AIAmazon Web Services (AWS)PythonVisual Studio

About

Experience

4 yrs 5 mos

Total Experience

1 yr 2 mos

Average Tenure

9 mos

Current Experience

American family insurance

Data Scientist

Sep 2025 – Present · 9 mos · Madison, Wisconsin, United States · Hybrid

Large Language Models (LLM)Retrieval-Augmented Generation (RAG)Python (Programming Language)Agentic AI Development

University of minnesota school of dentistry

GenAI Developer

Jun 2025 – Aug 2025 · 2 mos

American family insurance

Machine Learning Intern

May 2024 – Aug 2024 · 3 mos · United States · Remote

● Explored quantization frameworks like ONNX Runtime and OpenVINO to optimize YOLOx, YOLOv8, and Centernet models for odometer mileage extraction.
● Compressed model sizes to one-third of the original, maintaining 92-94% accuracy, and reduced evaluation runtime by 0.5 to 2 minute across 1,000 test images.
● Implemented static/dynamic quantization and quantization-aware training to convert models from float32 to int8 or uint8 format.
● Evaluated the accuracy of OCR and OCR_with_region APIs from Florence-2 VLM compared to YOLOx for odometer image processing.
● Fine-tuned a 2-layer LLaMA 3 model using LoRA and PEFT for efficient compression.

Neuro-Linguistic Programming (NLP)Generative AIProgramming LanguagesComputer VisionPython (Programming Language)Optical Character Recognition (OCR)+2

University of minnesota - carlson school of management

Graduate Teaching Assistant

Jan 2024 – Jun 2025 · 1 yr 5 mos · Minnesota, United States · On-site

● Teaching Assistant for "BA 3051: Data-Driven Business Decisions" at Carlson School of Management, responsible for creating exams/assignments, grading, and addressing student concerns in R and Excel.
● Guiding students through key concepts such as Linear Regression, A/B Testing, Time Series Analysis, and Optimization.

Microsoft ExcelStatistical Data AnalysisNeuro-Linguistic Programming (NLP)Programming LanguagesRjamovi+2

University of minnesota center for magnetic resonance research

Graduate Research Assistant

Jan 2024 – Dec 2024 · 11 mos · University of Minnesota - Twin Cities · Hybrid

● Worked on Motion correction for subsampled k-space MRI images of the brain. Utilizing score-based diffusion models and semi-supervised/unsupervised learning methods for motion correction.
● Exploring Image registration literature - studying multi-modal registration techniques for Deformable and rigid medical image registration.

Image ProcessingDiffusion Tensor ImagingProgramming LanguagesPyTorchC++

University of minnesota college of science and engineering

Graduate Student - AI and Robotics

Sep 2023 – Aug 2025 · 1 yr 11 mos · Minneapolis, Minnesota, United States

Convolutional Neural Networks (CNN)Neuro-Linguistic Programming (NLP)Programming LanguagesApplied Machine LearningGaNData Visualization+3

Credit suisse

2 roles

Quantitative Analyst

Jul 2022 – Aug 2023 · 1 yr 1 mo · Mumbai, Maharashtra, India · Hybrid

● Backgenerated zero rates for an RUB interest rate base curve switch (Mosprime to Ruonia) using an absolute move approach on historical Mosprime data, considering high correlation
● Debugging and enhancing of C++ and C# codebases for proper time series generation of specific interest rate curves and commodity derivatives
● Developing C# functions to reduce computation costs for generating curve objects

PythonStatistical Data AnalysisProgramming LanguagesVisual StudioQuantitative AnalyticsC+++2

Quantitative Analyst

Apr 2021 – May 2022 · 1 yr 1 mo · Mumbai, Maharashtra, India · Remote

● Validated an RNIV model methodology by Implementation testing - verified manual calculation with a C++ platform output
● Performed Portfolio level backtesting to check if PnL strips exceed the VaR+RNIV level, and Risk factor level backtesting to check no. of exceptions - when PnL exceeds the Expected Shortfall in the Stressed period.
● Regression testing- 'Variance Stabilizing Argument' was done to validate the optimum move methodology - proportional vs. absolute moves for risk factors - used for PnL strip generation with sensitivity data.

Microsoft ExcelPythonAnalytical SkillsProgramming LanguagesData AnalyticsHybrid Teamwork+1

Yatiken software technologies

NLP Intern

Jul 2020 – Aug 2020 · 1 mo · Delhi, India · Remote

● Conducted Multi-Lingual Sentiment Analysis using an RNN model for prediction of comment polarity.
● Utilized bidirectional LSTM, dropout layer for overfitting prevention, and Attention mechanism for sequence classification improvement.
● Achieved a model validation accuracy of 92%.

Programming LanguagesMachine Learning AlgorithmsAmazon Web Services (AWS)

Smiles.ai

Computer Vision

May 2020 – Jul 2020 · 2 mos · Bengaluru, Karnataka, India · Remote

● Built a real-time Head pose estimator using Dlib facial landmarks detector and OpenCV to capture images for tooth decay diagnosis.
● Predicted head tilt angle with 3D nose pointer and displayed alignment messages; demonstrated the same with opencv.js and tensorflow.js Posenet.
● Implemented CNN-based real-time smile detector with Haar Cascade Frontface Classifier for face detection.

Image ProcessingProgramming LanguagesMachine Learning AlgorithmsAmazon Web Services (AWS)

Salary dost

OCR and Deep Learning

Apr 2020 – Sep 2020 · 5 mos · Mumbai, Maharashtra, India · Remote

● Leveraged Google Cloud's Vision API and OpenCV for OCR on identity documents - achieved 85% accuracy. Developed Flask APIs for integration and trained an image classifier on GCloud for card type identification.
● Created an API for table extraction from salary slips (PDFs/images) using Pymupdf and Pytesseract, achieving around 70% accuracy.
● Developed an ANN model with 90% accuracy for predicting customers' loan returnability

Image ProcessingAnalytical SkillsProgramming LanguagesMachine Learning Algorithms

Department of mechanical engineering, iit kharagpu

Secretary at Mekanika

Aug 2018 – Apr 2019 · 8 mos

Event and publicity Coordinator; Organized tech competitions, workshops, industrial trips, department farewell and fresher’s welcome. Planned budgets and contacted several companies for sponsorship for events like "Goldberg Challenge" during technology fest ‘Kshitij’.

Programming LanguagesTeamwork