Shashank Padala
NewWinner, Health & Sciences Prize — Gemma 4 Good Challenge

Shashank Padala

I build and ship AI products — and write about how they actually get built.

Founder, Kirak Labs · ex-Amazon TPM II · ex-Tesco · IIT Guwahati · Queen's University

10+
Years Experience
10+
AI Products Launched
20M+
Enterprise Savings
3M+
Users Impacted
BACKGROUND

Education & experience

Education

Queen's University, Smith School of Business

Masters in Management, Innovation and Entrepreneurship

Indian Institute of Technology (IIT) Guwahati

Bachelor of Technology · Electronics & Communication Engineering

Experience

Founder · Kirak Labs Ltd (AI Dev Studio)

Apr 2025 – Present

Technical Program Manager II · Amazon

Sep 2022 – Apr 2025

TPM and Software Engineer · Tesco

Jul 2017 – Aug 2021

ACHIEVEMENTS

Awards & recognition

Hackathon wins, accelerator selections, and academic distinctions — 2017 to today.

2025

Selected Founder — AI Accelerator Programs

Up to $300K in credits

Google for Startups & Microsoft for Startups

Accepted into both accelerators for Kirak Labs, securing up to $150,000 in AI and cloud credits from each program.

2025

AI/ML and Vibe-Coding Workshop Host

Network School, Malaysia

Conceived, organized, and hosted workshops on Cursor, Claude Code, and Lovable for entrepreneurs and industry leaders at the startup ecosystem founded by Balaji Srinivasan.

2022

Winner — DAO Proposal Management

$5,000 grant

ETH Global Hackathon

Awarded a grant for a DAO-based project management tool. Also built CURE Data DAO at ETH Global — a decentralized platform aggregating encrypted hospital data on Filecoin, with on-chain access control via Medusa smart contracts.

2019

Winner — Garage Fest Internal Hackathon

+12% pick rate · ~£6M/year

Tesco

Recognized for redesigning retail picking-operations incentives. The resulting BRAG Bar lifted average pick rate 12% across 3,000+ stores, worth ~£6M a year.

2017

Winner — Internal Research Paper Competition

Tesco

"Efficient Retail Supply Chains Using Blockchain."

2017

Thesis Distinction

91% accuracy · 10/10

IIT Guwahati

One of a select group of B.Tech theses out of 600 final-year students — facial recognition in video, achieving 91% accuracy and a perfect 10/10 grade.

EXPLORING

Ideas & opportunities

What I'm spending thinking time on right now. None of these are products — each one says what stage it's at and what would have to happen next.

Open callVeterinary oncology · Personalized mRNA

Project Rosie, as a company

The hackathon pipeline works. The open question is whether it becomes a treatment rather than a demo: clinical-trial and compassionate-use partners, an order-of-magnitude cut to the ~$200–300K personalized dose, and distributed logistics for a molecule on a tumor's clock.

Looking for a technical co-founder and clinical partners.

See the pipeline
In progressMedical imaging · Multimodal ML

Achilles — knee MRI abnormality detection

RSNA AI Challenge on Kaggle: detect 12 clinically important knee MRI abnormalities from studies paired with their original radiology reports. Image plus text, scored on macro-AUC.

Submission deadline 22 Oct 2026.

EvaluatingDefense · Acoustic + passive RF

Heimdall — urban drone detection

A distributed acoustic and passive-RF mesh for detecting micro and mini UAS in cities, built to integrate into a broader C2 system. Written up as a PRFAQ, a product spec, and a system architecture — no hardware built.

Canadian IDEaS CUAS 2027 application closes 1 Dec 2026; demo in Ottawa, July 2027.

ConceptDefense manufacturing · Hardware

Project Avalon — sovereign fiber winding

Fiber-optic guided drones are the jam-proof answer to RF-jammed FPV, but rewinding raw telecom glass into high-speed center-payout aerial canisters is an outsourced chokepoint the West doesn't own. The thesis is re-shoring that one step in Ontario.

A one-page thesis. No facility, no LOI, no R&D validation yet.

IdeaGeospatial · Public safety

SALAMANDER — wildfire intelligence

Wildfire detection, intelligence, and response planning software. Drone swarm deliberately cut from the MVP — the interesting problem is the planning layer, not the aircraft.

Docs only. Not a commitment, and competing with better-validated lanes.

IdeaEducation · Personal brand

Build Your AI Portfolio

A course on shipping AI products, taught from 10+ real projects including the ones that died. The actual lesson is distribution — that's where most AI side projects end, not the model.

Unvalidated. Needs demand evidence before a single lesson gets written.

Shipped work lives in the portfolio. If any of these overlap what you're working on, I'd like to hear from you.

GITHUB

Open source & code

Recent public work. Full profile and history on GitHub.

github.com/shashank-padala

Loading repositories…

Shashank Padala
CONNECT

To connect with me

Reach out on LinkedIn — I respond to thoughtful messages about AI research, startups, partnerships, and advisory.

linkedin.com/in/shashank-padala
TRUSTED BY
AmazonTescoGoogle for StartupsMicrosoft for Startups
SYSTEM ONLINE