surabhi_pandey.py

~ /portfolio $ cat profile.py

Surabhi Pandey
1class SurabhiPandey:
2  role = "Healthcare AI · Biomedical Signals · Computer Vision"
3  affiliation = "IIT Guwahati"
4  # building interpretable, clinically reliable AI systems
5  def mission(self):
6    return "models that hold up under real clinical constraints"

Surabhi Pandey

01

about

{}

I'm Surabhi Pandey — a Data Sciences & AI student at IIT Guwahati focused on research-driven AI systems in healthcare, biomedical sensing, and computer vision. My interest centers on building models that are not only accurate, but clinically meaningful and technically rigorous.

I am particularly interested in medical imaging and wearable sensor data (IoT-based systems), and interpretable deep learning. I enjoy studying how models behave under real-world constraints such as patient variability, noisy signals, and domain shifts.

02

education

{}

B.Sc (Hons) in Data Science and Artificial Intelligence

Indian Institute of Technology, Guwahati

// 2023 – 2027

03

skills

{}
languages
PythonRMySQLHTML/CSS
ml_dl
Classical MLCNNsRNNsGANsVAEs
explainability
LIMESHAPGrad-CAMCounterfactuals
modeling_stats
Time SeriesBiomedical Signal ProcessingProbabilityModel Evaluation
tooling
NumPyPandasScikit-learnTensorFlowPyTorchMatplotlibSeabornGit
04

experience

[]
Machine Learning Intern — Hybionics Pvt. Ltd. // Feb 2026 – Present

Working on prosthetic IMU sensor signal modeling for motion analysis and fall-risk detection. Includes preprocessing, windowing, feature extraction, and robustness evaluation on real-world biomedical time-series data.

Research Intern — Prof. E.S.N. Raju P, IIT Guwahati // Jan 2026 – Mar 2026

Worked on explainable AI tools like LIME, SHAP, Grad-CAM, and counterfactual images on multi-type datasets. Worked on explaining black-box models such as ML models and CNNs.

Machine Learning Intern — BioScanAI Pvt. Ltd. // Oct 2025 – Jan 2026

Developed and evaluated machine learning models for retinal imaging–based disease classification. Performed data preprocessing, feature engineering, model validation, and performance benchmarking to improve robustness and predictive performance.

05

projects

[]

Explainable DL for Brain Tumor MRI

ResNet50Grad-CAMMRI

Trained a ResNet50 model for brain tumor classification (94% test accuracy) using patient-level splits. Integrated Grad-CAM and segmentation-guided counterfactual MRI generation to quantify causal dependence using Delta-Drop and Delta-Focus metrics.

Grad-CAM Repository Counterfactual Repository

OPD Record Management & Disease Prediction

FlaskMySQLFull-stack

Full-stack healthcare application with ML-based disease prediction. Built using Flask API, MySQL backend, and HTML/CSS frontend for managing patient records and symptom-based predictions.

View on GitHub

Stats Court

StatisticsWeb App

Web-based statistical testing system allowing CSV upload and automated hypothesis testing with structured analytical report generation. Supports classical statistical workflows with interpretive summaries.

View on GitHub View Website

ManoSakhi

Gemini APINLPMental Health

Gemini API-powered conversational system deployed on HuggingFace. Includes crisis keyword detection and intelligent response generation for supportive mental health interaction.

View on GitHub View Website

// for more, see my GitHub →

06

resume

{}

You can explore my full CV to see research projects, technical skills, and certifications.

download_cv.pdf
07

contact

{}
pandeysurabhi59@gmail.com linkedin/surabhi-pandey18 github/surabhipandey18
08

send_message

()