AI/Computer Vision & Data Science Intern

ASHISH KUMAR BISWAL

**************@*****.*** +916********* linkedin.com/in/ashish-kumar-biswal github.com/A-K-B-r EDUCATION

Indian Institute of Technology Madras – University of Birmingham July 2024 - January2026 Joint Masters in Data science and Artificial Intelligence GPA – 7/10, UK – Distinction Kalinga Institute of Industrial Technology,Bhubaneshwar September 2020 - May 2024 Bachelors in Computer Science and Engineering GPA – 8.38/10 WORK EXPERIENCE

Space Application Center,ISRO Ahmedabad, India

Research Intern August 2023 - December 2023

● Processed and analyzed data from 100+ INSAT-3DR satellite images to study the physical properties and spatial distribution of fog.

●Developed and evaluated a computer vision–based model for fog detection achieving an accuracy of over 90 percent while collaborating with 2 renowned scientists on the project. Highradius Bhubaneshwar, India

Summer Intern May 2023 - July 2023

●Worked on an AI-enabled B2B FinTech platform for automated invoice processing and accounts receivable optimization, directly relevant to MSME cash-flow management.

●Assisted in building data pipelines and ML-driven insights for credit risk assessment and payment prioritization. PROJECTS

SMART Wellbeing Platform (AI-Powered Full-Stack System) March 2026

●Built a scalable full-stack platform with role-based access (students, advisors, investors) and personalized dashboards, integrating a RAG-based AI chatbot using PostgreSQL and pgvector for context-aware responses.

●Optimized system performance by improving database query efficiency and implementing frontend optimizations (efficient data fetching, code-splitting), reducing load times and enhancing user experience. Segmentation of Clouds in Remote Sensing Data Using Deep Neural Architectures June 2025-August 2025

●Developed a novel multi-temporal U-Net for cloud segmentation on 2,700+ satellite image pairs, achieving 91% producer’s accuracy and reducing cloud omission errors.

●Engineered an end-to-end deep learning pipeline obtaining 79% balanced overall accuracy on Sentinel-2 data, improving discrimination between clouds and ambiguous land features. PUBLICATIONS

Automated Age & Gender Recognition in Networking Sites Using Optimized CNN Model February 2025

●Designed a CNN-based multi-face age and gender recognition system using OpenCV and a pretrained Caffe model, enabling simultaneous inference on multiple faces within a single image.

●Achieved 0.87 classification accuracy on real-world social media image samples, demonstrating robustness under unconstrained visual conditions.

Sustainable IoHT-Based ML Modeled System for Prediction of Cardiovascular Disease Risk August 2024

●Built an end-to-end ML pipeline integrating IoHT data with Logistic Regression, Decision Trees, and Random Forests for cardiovascular risk prediction.

●Performed feature analysis, model benchmarking, and ROC-based evaluation, identifying Random Forest as the best-performing model with 0.8836 accuracy

SKILLS

Programming Languages: Python, SQL, C, C++,Java

Competencies: Natural Language Processing, Large Language Models, A/B Testing, Regression, Gradient Boosted Machines, Text Processing, Hypothesis Testing,Computer Vision,Image Processing Tools/Frameworks: Git, Excel, Jupyter, AWS, Pandas, Numpy, Matplotlib, Seaborn, Scikit-Learn, Pytorch, Pyspark,Next.js, Django, PostgreSQL (pgvector), HuggingFace, PowerBI, Tableau ACHIEVEMENTS

●Secured 2nd place overall in BEAR Challenge – HPC competition at University of Birmingham.

Contact this candidate