08 · Final Round: Written Answers (verbatim)

info · Your text file 3c83bfd6-attachment.txt, kept exactly as written (spelling, "OPT", "kabootar" and all). The headings are your own bold headings, one per Round 3 question. The Whimsical links are yours too. The questions these answer are in 09 Final Round Whimsical Transcription (Part 1 and Part 2 of the Round 3 brief).

where does it live, how does someone actually start using it

ok so first thing, this is not an app. nobody is going to download a delivery otp app, thats basically a feature pretending to be a product. so it needs to sit inside places people already are everyday.

for the buyer, theres no app at all. its just a whatsapp number and a phone call, and the caller id literally shows "delhivery agentt." no download, no signup, no new habit. people already trust and check these channels all day.

for the delivery agent, it just becomes a screen inside the delivery console the delivery agent already opens like 50 times a day. not a separate app, just one more screen in the one he is already using everyday

for flipkart, its a simple toggle at checkout called "agentic cod" sitting right next to normal cod.

so basically three embed points, two existing apps (flipkart app and the ekart driver app) and one channel thats not even an app (whatsapp and voice calls).

https://whimsical.com/viraj-workspace3/final-round-madam-otp-5LZxz8Ln615eWX3bzdHspd@or4CdLRbgroLxWviWwKxLte2CxUKaR6NY4hjSuaJn

which company should build it, and why havent they already

flipkart, through ekart specifically. you need three things working together at once, delivery infra with live location, a payments rail that can hold and release money on its own, and a voice layer that can actually talk to someone in their language. flipkart is the only one in india with all three right now. ekart has delivery, super.money has the upi reserve pay stuff, and their SLAP team already has the voice muscle and they have patnered with RING ai.

the real question is why they havent built it already if they have all the pieces and I think , its more of an internal structure problem. EKART cares about cost per delivery and RTO numbers, not otp friction. super.money cares about transaction volume and app ranking, not whether a delivery actually lands. SLAP cares about shopping engagement, not what happens after checkout. three teams, three totally different goals, and this idea just falls into the gap between all of them where nobodys job depends on building it. https://whimsical.com/viraj-workspace3/final-round-madam-otp-5LZxz8Ln615eWX3bzdHspd@8ADn3nfZACaiRzMSXDtXMEZLoC7nK4e4Sj8h

who do they sell to, how do they make money

the buyer should not pay for this at the moment, the second it costs them anything the whole point falls apart. but there are three ways this makes money.

1) is just saving money internally and decreasing cost. RTO on COD orders in india is like 15 to 20% and every failed delivery costs real money, reattempts, reverse logistics, lost sales. every minute delivery agent is stuck at a door waiting for someone to read out an OPT is a minute he is not delivering the other 40 stops he has been assigned for. so when we measure the win through lower RTO and more deliveries per driver per day therefore increasing the overall throughput and increasing operational efficiency and optimisation.

2) (selling this to other D2C brand) ekart recently opened up its network to outside brands through a franchise setup. so this could become a paid addon inside that, like a "verified delivery" tier, priced as a small premium on top of regular cod, sold to brands who are already losing money to RTO and have no real way to fix it themselves.

3) they could license the whole thing, voice, geofence, payment capture, to other logistics companies who dont have all three pieces and sell at a premium amount.

what does everyone do today instead, and what does it cost them

. today its just someone reading an OPT out loud under pressure at the door, and for COD the buyer either pays cash or UPI but if he in unavailable then he has to call someone to pay to the delivery agent or just reschedule the next day which again cost him time and he has to plan this out. . the cost shows up as RTO, and wasted driver time.

and for non tech savy (househelp, secruity guard) its a very intimidating, tedious and Repetitive process

https://whimsical.com/viraj-workspace3/final-round-madam-otp-5LZxz8Ln615eWX3bzdHspd@or4CdLRbgroPTECnYu7MtD1J6Xn6YFCh6m6DnqLNZ

first 1,000 users

firstly this new behaviour of QR and the agenitc call (which will dynamically optimize the route by rescheduling) should be first taught to the deliery agent under ekart. they should know how to feature works because in the start, they might have to educate to the "buyer/receiver" and guide them in this process (manually doing this is much more effective than the voice agent calling and telling "we are sending you a QR")

i will pick one dense hub, say one ekart dispatch center in a single bangalore pincode cluster that drives high% of GMV . route the first 1,000 orders through just 15 to 20 delivery agentts at that one hub. keep it small and concentrated so we can actually sit and watch every call transcript and every geofence miss by hand in the first few weeks to debug any edge cases

and agentic COD Roll out on invite basis for cohorts with high frequency of COD and a known high RTO number - this is only for agentic payment for COD orders) they are core users and this feature can push them to power users cohort

where are these buyers already hanging out . inside the flipkart app at checkout, whatsapp messages from ekart. flipkart sathi aka flipkart's Whatsapp business chat name .

nothing really is getting someone to try it theres no opt in. its just the new default. the OTP step is gone, and since we are removing pain point and replacing it with just QR. this will probably get close to 90% percent adoption without spending much on marketing.

whats keeping them around after three weeks, for the buyer its just a few deliveries quietly going right, especially that moment where mom picks it up because the whatsapp message caught it early instead of a missed delivery and an annoying reschedule call. for arjun its about money and time, less standing around at doors, fewer failed drops he has to anways redo later.

for agentic COD , the user just turn on "agentic cod" that will be shown on flipkart account for that one pincode cluster. - this should be rolled out on invite only basis

https://whimsical.com/viraj-workspace3/final-round-madam-otp-5LZxz8Ln615eWX3bzdHspd@or4CdLRbgroLwtY1MKF66dr1h3FitSZGbjCAAHbq6

what could kill this in year one

assumption: lets assume the onboarding process of this agent(kabootar) through marketing and traning of delivery agent is done properly

regulation. RBi usually wants an extra factor of authentication for digital payments, and "geofence plus a photo equals approval" is exactly the kind of thing that might not meet that bar. if that gets flagged, the whole no touch payment idea goes back to needing a manual tap every single time, which kind of kills the whole point of building this in the first place.
quiet false negatives. if the voice agentts confidence scoring is even a little off, it starts marking people as unavailable who were actually home, which pushes rto up instead of down,
internal politics.. whoever owns this thing across teams can easily get deprioritized the moment theres a reorg or someones roadmap shifts, and a good pilot with no clear long term owner
token cost. gnani or whatever voice model this runs on charges per minute or per token for speech to text, text to speech, and the llm layer running underneath all of it. deliveries happen at a scale where even a tiny price increase gets multiplied across millions of calls a month, and we are talking two calls per order minimum, sometimes three if it escalates to arjun. if the voice provider raises prices, or if the confidence gate ends up triggering more calls than planned, the cost of running this per delivery could quietly end up higher than the savings it was supposed to create.

https://whimsical.com/viraj-workspace3/final-round-madam-otp-5LZxz8Ln615eWX3bzdHspd@or4CdLRbgroNthc5iu6GbxZGb5sKJkF918q26VKWp

note · The agent is kabootar (settled). The delivery agent in the story is Arjun; the buyer in the shoe story is Raj. Company: Flipkart through Ekart. Autonomy: L3.

Back to 00 README Index