Hi, I'm Akshay
💻 🎸 ☕️ 🧠 ❤️
25 | Data Science | Software Engineering | Artificial Intelligence
ContactAbout
My Introduction
With 4+ years of experience in AI, I design, develop, and deploy cutting-edge software and AI solutions, specializing in MedTech and beyond. I’ve led cross-functional teams, translating business goals into actionable Data & AI strategies that deliver results. Skilled in Python, C++, Java, and deep learning frameworks, I build scalable, high-performance systems. If you’re looking for someone to drive impactful projects, let’s connect!
Experience
My journey in the academic & professional frontB.Tech - Information Technology
Anna University, IndiaClass XII
Maths, Physics, Chemistry, Computer Science | SDAV Higher Secondary School, IndiaClass X
Central Board of Secondary Education | DAV School, IndiaLead MLE & Data Scientist
BrainSightAISenior MLE & Data Scientist
BrainSightAIMachine Learning Engineer
BrainSightAIProgrammer Analyst Trainee
CognizantArtificial Intelligence Engineer
DCKAPMachine Learning Lead Facilitator
Explore ML - Google AICore Developer
Google Developer Student Clubs - SJCECo-Founder
Pyxel AISkills
My technical & other skillsData Science & AI
4+ Years XPComputer Vision
95%Generative AI
95%Natural Language Processing
95%Signal Processing
90%Probability & Statistics
90%Data Analytics & Visualization
85%Frameworks & Libraries
90%Programming
3+ Years XPPython
95%SQL
95%C++/C
90%Java
90%Computing
3+ Years XPGPU & Distributed Computing
90%Amazon Web Services
85%Google Cloud Platform
85%Microsoft Azure
85%Front End
2+ Years XPHTML
90%CSS
85%JavaScript
75%React JS
85%React Native
85%BackEnd
3+ Years XPPython - Flask, Fast API
95%Firebase
75%Java - Spring Framework
75%Node JS, Express JS
70%Misc
4+ Years XPGit
90%Linux
90%Research
My research publications, conference contributions, & patentsAdvancing Pediatric Brain Mapping: An AI-Driven Adaptive Functional MRI Pipeline for Mapping of Functional Networks
Conference: Radiological Society of North America (RSNA), 2024Link: To be updated
Author(s): Akshay Kumaar M, Sachin Patalasingh, Malvika Ganesh, Radha Kumari, Rimjhim Agrawal
A self-supervised learning based adaptive pipeline for identification of paediatric functional brain networks.
Accelerated Motion Correction with Deep Learning for Functional MRI - A potential for real-time fMRI processing for faster clinical assessment
Conference: Radiological Society of North America (RSNA), 2024Link: To be updated
Author(s): Akshay Kumaar, Saurabh Jain, Sachin Patalasingh, Shamanth Hampali, Janova Anbarasi, Rimjhim Agrawal
A deep learning based accelerated motion correction pipeline for fMRI data using a Dual-Head ResNet Regressor.
Harmonizing tb-fMRI and rs-fMRI: A Generative approach for mapping Language Networks
Conference: European Congress of Radiology, 2024Link: EPOS ECR 2024
Author(s): Akshay Kumaar M, Sachin Patalasingh, Rimjhim Agrawal, Saurabh Jain, Dr. Sunita P Kumaran, Dr. Shreyas Reddy Kankara, Dr. Narayan Menon, Dr. Shailesh. A. Rao
Utilizing generative modelling to map task-based fMRI analogous brain activity maps for Language network using rs-fMRI connectivity.
Brain Tumor Classification using a Pre-Trained Auxiliary Classifying Style-based Generative Adversarial Network
Available at: IJIMAI | Progress: PublishedLink: https://doi.org/10.9781/ijimai.2023.02.008
Author(s): Akshay Kumaar M, Dr. Duraimurugan Samiayya, Venkatesan Rajinikanth, Durai Raj Vincent P M, Seifedine Kadry
An auxiliary classifying conditional generative adversarial network based on StyleGAN, achieving ~99.5 % accuracy in classifying Glioma, Meningioma, & Pituitary Brain Tumors from MR Images.
Does resting-state fMRI have the potential for Presurgical functional Mapping?
Conference: Radiological Society of North America (RSNA), 2023Link: To be updated
Author(s): Akshay Kumaar M, Sachin Patalasingh, Dr. Radha Kumari, Malvika Ganesh, Janova Anbarasi, Ruchi Sharma, Dr. Rimjhim Agrawal
Exploring the utility of resting-state fMRI for presurgical functional mapping in comparison to task-based fMRI.
A Novel Connectomic Analysis Framework for Personalized Presurgical Planning using Structural MRI, rs-fMRI, and DWI
Conference: Radiological Society of North America (RSNA), 2023Link: To be updated
Author(s): Saurabh Jain, Nayan Wadhwani, Akshay Kumaar M, Dr. Rimjhim Agrawal, Ruchi Sharma, Malvika Ganesh Iyer, Dr. Sarbesh Tiwari
A novel framework for personalized presurgical planning using structural MRI, rs-fMRI, and DWI data for brain tumor patients.
Deep Learning based head motion estimation for correction between volume motion artifacts in diffusion MR images
Conference: Nature Conference (2022) Neuroimaging in AI, 2022Link: To be updated
Author(s): Shamanth Hampali, Akshay Kumaar M, Saurabh Jain, Ranganayaki Sathyanarayanan, Rimjhim Agrawal
A deep learning based head motion estimation pipeline for correction of motion artifacts in diffusion MRI images using Convolutional Neural Networks.
VoxelBox: A Novel Tool for Dynamic Network Localization of Brain Functional Activity in Dementia using rs-fMRI
Conference: Federation of European Neuroscience (FENS), 2022Link: To be updated
Author(s): Dr. Rimjhim Agrawal, Akshay Kumaar M, Dilip Rajeswari, Ruchi Sharma, Sahana Hegde, Laina Emmanuel, Ranganayaki Sathyanarayanan
A novel tool for dynamic network localization of brain functional activity in dementia using resting-state fMRI data and comparison with healthy controls.
A Hybrid Framework for Intrusion Detection in Healthcare Systems using Deep Learning
Journal: Frontiers in Public Health | Progress: PublishedLink: 10.3389/fpubh.2021.824898
Author(s): Akshay Kumaar M, Dr. Duraimurugan Samiayya, Durai Raj Vincent P M, Kathiravan Srinivasan, Chuan-Yu Chang*, & Harish Ganesh
A Machine Learning based approach to detect network & application intrusions, achieving ~99.2% accuracy in successfully detecting around 19 types of latest cyber attacks.
System and Method for determining Networks of Brain from Resting State MRI data using ML
Patent Number: US20240366087A1 | Progress: IssuedLink: US Patent Center
Author(s): Dr. Rimjhim Agrawal, Ruchi Sharma, Dilip Rajeswari, Akshay Kumaar Murali Gopika Manoharan
A method for determining brain networks using resting-state MRI data and machine learning to evaluate a subject's health. It involves processing MRI data, converting it to object format, generating a 4D connectivity file, and analyzing brain networks with a machine learning model.
Projects
My independent projects & contributions
Neural Black - AI for Brain Tumor
Neural Black is a complete brain tumor detection, & classification system with high accuracy that uses CNNs.
Loading...
Drive AI - Warning System for Drivers
Uses Computer Vision & Deep Learning that is capable of detecting traffic signs & signals, driver drowsiness, lane changes, pedestrian movements, & sensor features (similar to autonomous driving) which are analyzed by a sophisticated system to warn the driver.
Loading...
Third Eye - Intrusion Detection System
A Hybrid Intrusion Detection system with firewall that uses Machine Learning to detect network intrusions by analyzing network traffic data and extra environmental features.
Loading...
Image Super-Resolution GAN
Generating super-resolution images from low-resolution images using Super Resolution Generative Adversarial Network that looks photo-realistic.
Loading...
Malware Prediction
Malware Prediction for portable executable applications using Machine Learning to predict if an executable program is a malware or is in a risk to get attacked
Loading...
Stock Bot
A simple stock bot that uses Reinforcement Learning (Deep Q Learning) to sell, hold or buy stocks by taking decisions automatically. Simulates profit and loss based on stock investments
Loading...
Tofu - Deep Learning Framework
Tofu is a Deep Learning Framework for Python, C, C++, and Java that aims to deliver high user-friendliness and user-experience simplifying technicality and improving accessibility to the general population.
Loading...