Experience

A collection of things I've figured out, built, and broken.

Research across RNA therapeutics, single-cell immunology, machine learning and plant biotechnology.

Scroll through the roles to see what I worked on and what came from each project.

Wet-lab questions, computational experiments, and the occasional side quest—collected in one place.

Roles06

Institutions04

Timeline2024–2026

01Jun 2026 – Sep 2026 · 4 months

National Centre for Biological Sciences (NCBS), KoshKey

41 RNA mechanisms, one platform. Submitted to ICLR 2027.

Bioinformatics Scientist · Supervised by Dr. Tanay Bhatt

Built a platform covering 41 RNA therapeutic mechanisms and 9 therapeutic goals, using 19 mechanism-specific features to rank candidate oligos.
Compared 7 ML models across 12 controlled experiments, then checked the workflow with 84 arbitration tests and 11 feature-level tests. The work was also sent for patent.
Used 10x Genomics scRNA-seq to study mouse fibroblast heterogeneity in scleroderma and helped design primers for follow-up experiments.

ASO / siRNA design
ML benchmarking
10x Genomics
Seurat
Monocle3
CellPhoneDB

02Oct 2025 – Mar 2026 · 6 months

Dept. of Biochemistry, Indian Institute of Science (IISc)

Built HPC workflows for stimulated T-cell cohorts.

Research Student · Advised by Prof. Dipankar Nandi; co-advised by Dr. Payel Roy

Used CellRanger, FeatureCounts and Salmon-Alevin to process stimulation-resolved cardiovascular and immune cohorts, with Harmony/scVI batch correction and Scrublet for doublet removal.
Identified stimulated T-cell signatures with limma-voom, MAST and GLMMs, profiled TCR/BCR clonotypes with MiXCR, and built repeatable SLURM workflows with Snakemake, Nextflow and Docker.

CellRanger
Harmony / scVI
limma-voom
MiXCR
Snakemake
Nextflow

03Jun 2025 – Jun 2026 · 13 months

Indian Institute of Science (IISc)

Awarded the Research Excellence Honour.

Project Trainee, B.Tech Internship Dissertation · Advised by Dr. Payel Roy

Studied bulk and single-cell RNA-seq data from human and mouse T-cell populations in chronic atherosclerosis.
Built a reproducible pipeline for quality control, differential expression, cellular heterogeneity, trajectory and pseudotime analysis, and Th17 metabolic modelling.
The manuscript was submitted to the Journal of Immunology (American Association of Immunologists).

RNA-seq
STAR
DESeq2
Seurat
WGCNA
COMPASS

04Nov 2024 – Jul 2025 · 9 months

Dept. of Computer Science, PES University

First-authored at NeurIPS 2025 workshops.

Research Student · Advised by Prof. Sunil K C; co-advised by Prof. Badri Prasad V R

Built TriggerNet, an explainable AI framework that combines Grad-CAM, RISE, FullGrad and TCAV in a hierarchical pipeline. It reached 97.3% accuracy on plant images, 94.2% on PathMNIST and 92.8% on CIFAR-10.
The work was recognised among the top papers at SPIGM and Reliable ML from Unreliable Data, and was presented at ICCET. It improved the p-score by 21.4% and lowered the Brier score by 16.7% compared with the baselines.

Explainable AI
Grad-CAM
RISE
FullGrad
TCAV

05Jul 2024 – Dec 2024 · 6 months

Center for Innovation and Entrepreneurship (CIE)

One of 120 chosen from 1,500 applicants.

Research Student · Advised by Dr. Kavitha S H; co-advised by Prof. Sathya Prasad

Designed Kavach, a platform for detecting manipulated content, and built MATLAB models of S. cerevisiae bioethanol production across several feedstocks using more than 250 sources.
Optimised A. awamori cellulase production with Plackett–Burman and response-surface methods. The model reached R² 0.9878, with 87% of peak titre in 58% of the time.

Deepfake detection
MATLAB
Kinetic modelling
Plackett–Burman
RSM

06Feb 2024 – Jun 2025 · 17 months

Plant Tissue Culture Lab, PES University

Honours thesis in the top 10% of the cohort.

Research Intern · Advised by Dr. Reshma S V

Worked on plant tissue culture, inoculation and hardening, as well as quorum-quenching assays and SEM on P. aeruginosa.
Ran virtual screening, docking and GROMACS molecular dynamics with MMPBSA to study phytochemicals against QS targets.
Outcome: Honours thesis integrating in vitro and in silico work, top 10% of the graduating cohort.