🏆 Secured 1st position in our Problem Statement at the Smart India Hackathon 2025 Grand Finale. 🚀
Most teams race against the clock.
We treated time as a constraint to design around, not a threat to react to.
36 hours. ⏰
6 team members.
1 problem statement.
That was the only framework that mattered.
Pressure does not break teams.
Lack of structure does.
We focused on execution discipline rather than urgency, operating with clearly defined roles (which blurred as time went by). The final hour was structured and deliberate (and very chaotic) — narrative refinement, prototype stress-testing, and end-to-end validation running in parallel.
There were no individual heroics (well, maybe there was) — only deliberate division of responsibility.
We intentionally chose a high-complexity, high-impact problem statement proposed by MathWorks:
"AI-powered monitoring of crop health, soil conditions, and pest risks using multispectral/hyperspectral imaging and sensor data."
The challenge addressed real constraints in Indian agriculture — soil degradation, unpredictable weather, and pest outbreaks — where traditional monitoring remains delayed and imprecise.
Our solution focused on simplicity by design: a platform that converts complex spectral and sensor data into actionable insights, forecasts, and localized alerts for agronomists, researchers, and farmers — without requiring deep expertise in remote sensing.
What set it apart wasn’t feature count, but focus.
Clear signals.
Timely intervention.
Decisions accessible without specialized technical knowledge.
The journey was not linear. We spent more time than expected on time-series modeling and integration, and API credit limits forced hard trade-offs (In fact, it expired again when I was taking screenshots for this post). These constraints sharpened prioritization and reinforced architectural discipline under pressure.
During evaluation, the judges emphasized the cautious and responsible use of Large Language Models, highlighting the importance of validation, reliability, and domain grounding in high-stakes applications.
Grateful to the team — Aditya Chauhan, Achyuth A, Adith Jayakrishnan J, Dhyan Shah, and Priti Nag — for the discipline and ownership throughout the sprint.
Sincere thanks to our mentor Dipankar Mandal Sir, our judges Ramana Reddy A V Sir, Abhishek Tripathi Sir, and Sourabh Sathe Sir, and to AICTE Vice Chairman Dr. Abhay Jere for making innovation of this scale possible. We would also like to thank all those seniors whom we disturbed for the entire duration of SIH for reviewing our slides, prototype and in general their advice.
#SmartIndiaHackathon #SIH2025 #AgriTech #PrecisionAgriculture #AIForGood #MathWorks #MachineLearning #ResponsibleAI #Teamwork
18 Décembre 2025 à 08h18
Voir sur LinkedIn →
