Open to opportunities

Kanishka S J

@kanishkasj

Message

AI & Data Science undergraduate skilled in deploying production-ready computer vision and ML systems.

India

Message

What I'm looking for

I'm seeking an entry-level ML/AI or computer vision role where I can deploy production models, work with embedded systems and cross-functional teams, and continue learning while delivering measurable real-world impact.

I am an AI & Data Science undergraduate with hands-on experience deploying high‑accuracy machine learning models into real‑world environments, specializing in computer vision and full‑stack development.

I designed and deployed an autonomous Sentinel Turret Rover using ArcFace with CA‑CBAM-enhanced CNN for face recognition, achieving 95% testing accuracy and real‑time alerting via Telegram Bot on Raspberry Pi and Arduino hardware.

I developed an AI drone system for herbal leaf classification using ResNet‑50 and Vision Transformers with 97.68% accuracy, and built a 5G packet‑loss predictor using LSTM/CNN that reached 99.63% accuracy to improve network QoS.

I hold certifications in web development, generative AI, and deep learning, earned a Problem Solver Award at HACKSKYE 2025, and I thrive on building production-ready ML solutions that deliver measurable impact.

Experience

Work history, roles, and key accomplishments

Sentinel Turrent Rover

Amrita School of Engineering

Apr 2025 - Present (1 year 2 months)

Designed and deployed an autonomous rover-based turret using ArcFace with CA-CBAM-enhanced CNN for intruder detection and precision targeting, achieving 95% test accuracy and real-time Telegram alerting.

Python CNN ArcFace Raspberry Pi Arduino Computer Vision Real Time Systems

Herbal Leaf Classification

Amrita School of Engineering

Nov 2024 - Present (1 year 7 months)

Built an AI-powered drone system using ResNet-50 and Vision Transformers for real-time herbal leaf classification, achieving 97.68% accuracy with low-latency onboard inference and geo-tagged alerts.

ResNet50 ResNet 50 Vision Transformer Python Onboard Inference Computer Vision

5G Packet Loss Prediction

Amrita School of Engineering

Apr 2024 - Present (2 years 2 months)

Developed an LSTM-CNN model to predict packet loss in 5G mmWave networks and classify LoS/DNLoS/SNLoS, achieving 99.63% accuracy to enhance QoS.

CNN TensorFlow Time Series Analysis MmWave Model Evaluation

Lead Adsorption Optimization

Amrita School of Engineering

Apr 2024 - Present (2 years 2 months)

Developed SVR models with Bayesian optimization and bootstrapping to predict and optimize Pb adsorption, achieving a mean absolute error of 8.22 and identifying key influencing factors.

scikit learn Bayesian Optimization Bootstrapping Data Analysis Python Experimental Design

Education

Degrees, certifications, and relevant coursework

Amrita School of Engineering

Bachelor of Technology, Artificial Intelligence & Data Science

Grade: 8.96/10

Pursuing Bachelor of Technology in Artificial Intelligence & Data Science at Amrita School of Engineering with a GPA of 8.96/10 and coursework in Gen AI, Machine Learning, Deep Learning, Computer Vision, Transformers, DSA, and OOP. Expected May 2027.