The Rain Doesn't Care About His EMI

A story about Karthik, 15 million invisible workers, and why we built RiderNet.

ACT 1: His World
It is 11:45 AM on a Wednesday in Sathyamangalam, Erode.
Karthik is already checking his helmet strap before he's fully out the door.
He does this every day — the same routine, the same sequence. Wake up. Make tea. Drop his kids at school. Come back. Eat whatever is left on the stove. Check the Swiggy app. Strap the helmet. Go.
Today feels important. School fees are due Friday. His wife mentioned it last night. She mentioned it again this morning. He didn't need the reminder — he'd been calculating in his head since Sunday. If he hits 18 to 20 deliveries today and tomorrow, they'll be fine. Just about fine.
He looks at the sky before starting his bike.
Grey. But manageable.
He leaves.

By 12:15 PM, he's near the Sathyamangalam town area, waiting outside a restaurant for his first order of the afternoon. The lunch rush is beginning. This is the golden window — 12 PM to 3 PM — where a good rider can stack deliveries back to back and make the kind of money that actually matters. Karthik knows these streets better than most. He knows which roads to take when orders stack up, which shortcuts save three minutes, which restaurant gets busy first.
He knows everything about how to maximize a good day.
What he doesn't know — what nobody told him — is how to survive a bad one.
At 12:28 PM, the rain starts.
Not a warning. Not a drizzle that gives you time to think. Just — rain. Hard, immediate, the kind that turns Sathyamangalam's streets into shallow rivers within twenty minutes. The kind that makes the roads he knows by heart suddenly dangerous. The kind that doesn't care about school fees or bike EMIs or the stack of orders waiting on his app.
Karthik pulls over near a tea stall on the main road.
He turns off the engine.
And he waits.

Around him, the lunch rush he had planned his entire week around is happening without him. Somewhere, orders are being placed. Somewhere, other riders who managed to leave earlier are completing deliveries. Somewhere, his earnings counter on the Swiggy app sits completely still.
He checks his phone. One order available — 4.2 km away, through the flooded stretch near the bus stand. He stares at it for a moment.
His bike is under EMI. ₹3,200 every month, due on the 5th. If the bike breaks down in a flood — and it has happened before, to riders he knows — he doesn't just lose today's earnings. He loses the ability to earn at all until it's fixed. And repairs cost money he doesn't have sitting around.
He declines the order.
He sits at that tea stall for two hours and eleven minutes.
When the rain finally eases enough to ride safely, the lunch rush is over. The golden window is gone. He manages four deliveries between 3:15 PM and 5 PM. Four deliveries after planning for twenty.
He earned ₹140 that afternoon.
That evening, he calls his wife from the parking area outside a restaurant.
The school fees will be late.

ACT 2: The Broken System
Karthik is not unlucky. Karthik is unprotected.
There is a difference — and it matters enormously.
There are more than 15 million gig delivery workers in India today. They are the engine behind Swiggy, Zomato, Blinkit, Zepto, and Amazon. Every hot meal that arrives at your door, every grocery order that beats your expectations — there is a Karthik behind it. Riding through heat, rain, traffic, and exhaustion, operating on a simple and brutal equation: work today, earn today, survive today.
When external disruptions hit — and they hit constantly — the math collapses.
Heavy monsoon rains. Extreme heat waves. AQI alerts that make outdoor air dangerous to breathe for hours at a time. Sudden bandhs with no warning. Flash floods that turn familiar roads into obstacles. These aren't rare events in India. For a delivery rider, these are regular features of the working year — and every single one of them translates directly into lost income with zero recourse.
We looked for existing solutions. We looked hard.
Traditional insurance doesn't work for Karthik. Not even close. Traditional insurance requires proof of loss — documents, receipts, an adjuster's visit, a claims process that takes weeks. Karthik doesn't have payslips. He doesn't have an employment contract. He can't prove his income on paper because no paper exists. And even if the claim were eventually approved, what use is a payout three weeks later to a man whose family needed money on Friday?
Government schemes exist — but they are designed for formal workers, not platform gig workers who exist in the grey space between employment and entrepreneurship.
The platforms themselves offer nothing for income lost to weather.
So Karthik sits at that tea stall. And he absorbs the loss alone. Every time.
This is not a small inconvenience. For a sole breadwinner with a bike EMI, young children, and zero savings buffer — one bad rainy afternoon can trigger a cascade. Late school fees. Skipped groceries. Borrowed money from relatives. Stress that doesn't leave when the rain does.
And nobody — not a single system, institution, or product — was built to catch him.
Until now.

ACT 3: RiderNet Changes Everything
We are a team of students — but we are also people who grew up in cities like Erode, who have seen Karthik a hundred times without knowing his name. The rider waiting outside a restaurant in the rain. The delivery partner who smiles and hands you your order even when his shoes are soaked. The man on the bike whose entire livelihood depends on whether the sky cooperates today.
We didn't build RiderNet to win a competition. We built it because we couldn't look away from a problem that was right in front of us — real, urgent, and completely unsolved. Karthik's story is not an edge case. It is the daily reality of 15 million workers who power India's digital economy while the system looks the other way.
We chose to look directly at it. And then we chose to build.

RiderNet is an AI-powered parametric insurance platform built exclusively for food delivery workers.
The word parametric is everything. Let us explain it the way we explained it to ourselves when we first understood it.
Traditional insurance asks: "Can you prove you lost income?"
Parametric insurance asks one different, simpler, more powerful question: "Did it rain more than 64.4mm in Sathyamangalam today?"
If the answer is yes — confirmed by OpenWeatherMap, verified against IMD data, cross-referenced with Karthik's registered delivery zone — then the payout triggers. Automatically. Instantly. Without Karthik filing a single form, making a single call, or proving a single thing.
The rain is the proof. The data is the claim. The payout is immediate.

Now imagine Karthik's Wednesday with RiderNet.
Same rain. Same flooded streets. Same dangerous roads near the bus stand.
But at 12:35 PM — seven minutes after the rainfall threshold crosses the trigger point — Karthik's phone buzzes with a notification:
"🌧️ Heavy rainfall detected in your delivery zone — Sathyamangalam, Erode. Parametric trigger activated. Income protection of ₹600 has been credited to your UPI account."
He didn't file anything. He didn't call anyone. He didn't prove anything.
He just received the safety net that should have always existed.
His kids' school fees are paid on Friday. Not late. On time.

How does RiderNet know when to trigger?
We built five parametric triggers — each one representing a real, verifiable, zone-wide disruption that causes genuine income loss for delivery riders:
T1 — Heavy Rainfall when precipitation crosses 64.4mm in 24 hours, the IMD's official red alert threshold. Verified through OpenWeatherMap API.
T2 — Extreme Heat when temperatures exceed 45°C and a government heat wave declaration is active. Because riding in dangerous heat isn't just uncomfortable — it's income-destroying.
T3 — Severe Air Pollution when AQI crosses 400 (Severe+ category) triggering GRAP Stage IV restrictions. Verified through CPCB and OpenAQ data.
T4 — Flooding and Waterlogging when specific zones report road flooding beyond rideable conditions. Verified through government alert data and geo-fenced zone monitoring.
T5 — Curfew or Local Strike when a verified bandh or zone closure makes pickup and drop locations inaccessible. Verified through News API and geo-fence triggers.
Every trigger is external. Every trigger is verifiable. No single rider can fake any of them — because the data comes from sources entirely outside their control. And when a trigger fires, it fires for every rider in the affected zone simultaneously — which is itself proof that the disruption was real.

How much does it cost Karthik?
Every Monday morning, RiderNet calculates Karthik's personalized weekly premium using our AI engine — built on XGBoost, trained on historical disruption data, weather patterns, and zone-level risk profiles.
The premium ranges from ₹29 to ₹79 per week.
To put that in perspective — ₹29 is less than a single cup of tea and a biscuit at that tea stall where Karthik waited in the rain. For that price, his entire afternoon earning window is protected against weather-driven income loss.
The premium is weekly because Karthik thinks weekly, earns weekly, and plans weekly. A monthly premium doesn't fit his life. A weekly one does — it comes out of this week's earnings and protects this week's income. Simple, aligned, fair.

How we are building it
We are building RiderNet on a React and FastAPI stack, with PostgreSQL for data storage and XGBoost powering our dynamic premium engine. Our parametric triggers connect to OpenWeatherMap, OpenAQ, and CPCB APIs in real time. Fraud detection runs on an Isolation Forest model that validates GPS location, activity patterns, and cross-references claims against verified trigger data. Payouts are processed through Razorpay's test mode — simulating instant UPI transfers to the worker's registered account.
We are not just describing an idea. We are building the actual system, week by week, line by line.

The Closing: Karthik's Friday
It is Friday morning in Sathyamangalam.
Karthik drops his kids at school. His daughter runs in without looking back — she does that now, confident, comfortable, knowing everything is okay.
The fees were paid yesterday.
He gets back on his bike, checks the sky — clear today — and opens the Swiggy app. Twelve orders already queued for the afternoon rush. He adjusts his helmet, starts the engine, and pulls out onto the road.
Somewhere in the background, RiderNet is already running. Checking the forecast. Calculating next Monday's premium. Monitoring his zone. Ready — silently, automatically — to catch him if the sky turns grey again.
He doesn't think about it. He doesn't have to.
That's exactly the point.

There are 15 million Karthiks across India. Riders who absorb every storm alone. Workers who power an economy that has never powered them back.
We built RiderNet because the technology to protect them has always existed. The parametric triggers, the AI models, the instant payment rails — none of this is new. What was missing was someone deciding that delivery riders were worth building it for.
We decided they were.
Because when it rains in Sathyamangalam, it shouldn't just be Karthik's problem.

We are Squad 4, building RiderNet at Guidewire DEVTrails 2026 University Hackathon. Follow our journey as we build, iterate, and fight to put a safety net beneath every rider in India.
If this story resonated with you — share it. Every Swiggy order you've ever placed had a Karthik behind it.