Jonathan Dorairaj

Hi, I'm Jonathan Dorairaj

Complexity is fun.


Introduction

About Me.

Developer turned Machine Learning Engineer, I tailor and implement data driven solutions to help breakthrough domains. My passion of maths backing ML/AI helps me design solutions that are just right.

Machine Learning & AI

Data Science & Analysis

Cloud & Engineering

Python Developer

My Work

Projects.

Parameter Efficient Studies of Large Language Models

Explored parameter-efficient fine-tuning (PEFT) methods for large language models, focusing on LoRA
and Adaptors. Introduced a novel application of Laplace Approximations on adaptor layers to improve
model calibration and uncertainty estimation.

Exploring Uncertainty in scDEAL: A Variational Autoencoder Perspective

Transitioned a drug sensitivity prediction model from a deterministic to a probabilistic framework usingVariational Autoencoders (VAEs), leveraging both bulk and single-cell expression data via transfer learn-
ing.

Role-Based Access Control Chatbot Using RAG

FastAPI backend with RBAC, Claude and Chroma semantic search

Tempest: Migraine Tracker Google Extension

Chrome Web Extension for tracking and logging migraine events. Integrated with Google API.


My Journey

Work Experience.

Research Assistant - Digital Twins, Karolinska Institute

February 2025 – Present

• Analyzed single-cell RNA sequencing data using Python and R for biomarker identification and target
discovery
• Developed digital twin models and disease trajectory analyses for early diagnosis and prevention
• Constructed multicellular models and applied pseudotime approaches to map cellular differentiation and
disease progression
• Applied machine learning algorithms to proteomics datasets for pattern recognition and biomarker vali-
dation
• Built multicellular communication models to analyze cell-cell interactions

Senior Information Security Analyst, Wipro

May 2021 – Jan 2022

• Improved time to resolution by 10-15% team-wide by implementing a dashboard
• Improved playbook response time by 30% by troubleshooting Cortex SOAR incidents
• Awarded 3 times for individual contributions.
• Top 10% in the team by performance metrics.
• Investigated 200+ network attacks and malicious files, effectively quarantining and sanitizing affected
endpoints.
• Investigated over 1000+ phishing and malware incidents

Information Security Analyst, Wipro

Jan 2019 – April 2021

• Investigated and resolved over 7000+ alerts, events, and incidents.
• Ranked in the top 10% for productivity, delivering the highest performance metrics in the team.
• Identified and blocked over 200+ Indicators of Compromise.
• Monitored network for alerts and DDoS attacks with 100% record.


send me a message

Contact.