I wish you could bet on the odds of some of these succeeding in 3 years or so. Someone should start a startup that does this and get into Y Combinator to be super meta.
random commentary below:
* FloydHub – Heroku for deep learning *
It's unlikely in my eyes that a company with enough data worth doing deep learning for is going to outsource the actual model building. I'd be interested to hear counter examples.
* Tetra – Automatic notes for business meetings *
I think this is worthless. Meetings important enough to warrant for "paid notes" will have people taking notes, on their own (company) dime anyway. Not to mention taking notes will improve your engagement in the meeting. Not to mention the privacy implications.
* Collectly – Stripe for medical debt collection *
Pretty interesting. Hard to comment without knowing what the success rate is for an average debt collector. I fear there's some selection bias in the stats techcrunch shared.
* Indigo Fair – Amazon for local retailers*
Unfortunately I think local retailers are slowly dying. I do think this is an interesting niche that could be successful.
* Lively – Modern healthcare savings account (HSA)*
This too could be successful. Most people who'd do this probably have an HSA that's offered by their employer, unless they're trying to replace that.
* KidPass – One pass for “amazing activities for kids”*
I'm surprised at the success this is seeing. I do think parents are a good group to exploit though. Overworked parents want the logistics of spending time with their kids to be outsourced these days. Plenty of rich parents in NYC too.
* Upcall – Outbound calls as a service*
Not a huge fan of spamming services.
* Niles – Conversational wiki for business *
It's cheap enough, I guess.
* Sycamore – Onboarding drivers for on-demand jobs *
Looks interesting, but don't see how it'll address the drivers making so little part techcrunch mentioned.
Upcall is definitely not a spamming service (no robotcalls or autodialing service), but a human-powered platform supplementing SDR and CSR at companies looking to scale their outbound leads efforts and customer retention/engagement. Upcall is a SaaS/API which connects clients with work from home U.S. based callers, making calls within a specific context (qualify lead, welcome user on a platform, collect feedback, reactivate user, etc...).
> Upcall is definitely not a spamming service (no robotcalls or autodialing service)
Nobody said it was a robot-calling or auto-dialing service.
"Qualifying leads" is the thing where people with whom I have no prior business arrangement call me at work, asking if I'm the right person to talk to about (service that they sell), right? "Spamming service" sounds about right.
I had the idea for upcall several years ago around the time a similar enterprise service that went through YC (elasticsales?) was starting up. So I am biased but it is a good idea. Sales and marketing is about targeting the appropriate customer so no one on either end wants to waste time. Spam would be calling everyone and hope someone buys what you are selling. So upcall is great for B2B startups without a budget for sales hires. (Maybe startups will also focus on sales from the beginning rather than shying away from it like VCs complain about.)
We (skymind w16) build dl models on prem for customers after licensing our deployment software. We sell to fintech,telco, retail, car makers, plane makers and device makers. On prem and connectors to sap turned out to work.
Flynn will be great for education and proof of concepts.
Just realized misspelled. Floyd* Anyways - it's a great model that can work, but the "cloud api" has been tried to death. See: most recent acquisitions. Maybe things will change, it'll take time though.
> Someone should start a startup that does this and get into Y Combinator to be super meta.
Not worth a startup, but a fantasy-sport-like site that allowed some "gambling" -- perhaps "paid" for with karma points -- on YC startups would be fun.
random commentary below:
* FloydHub – Heroku for deep learning *
It's unlikely in my eyes that a company with enough data worth doing deep learning for is going to outsource the actual model building. I'd be interested to hear counter examples.
* Tetra – Automatic notes for business meetings *
I think this is worthless. Meetings important enough to warrant for "paid notes" will have people taking notes, on their own (company) dime anyway. Not to mention taking notes will improve your engagement in the meeting. Not to mention the privacy implications.
* Collectly – Stripe for medical debt collection *
Pretty interesting. Hard to comment without knowing what the success rate is for an average debt collector. I fear there's some selection bias in the stats techcrunch shared.
* Indigo Fair – Amazon for local retailers*
Unfortunately I think local retailers are slowly dying. I do think this is an interesting niche that could be successful.
* Lively – Modern healthcare savings account (HSA)*
This too could be successful. Most people who'd do this probably have an HSA that's offered by their employer, unless they're trying to replace that.
* KidPass – One pass for “amazing activities for kids”*
I'm surprised at the success this is seeing. I do think parents are a good group to exploit though. Overworked parents want the logistics of spending time with their kids to be outsourced these days. Plenty of rich parents in NYC too.
* Upcall – Outbound calls as a service*
Not a huge fan of spamming services.
* Niles – Conversational wiki for business *
It's cheap enough, I guess.
* Sycamore – Onboarding drivers for on-demand jobs *
Looks interesting, but don't see how it'll address the drivers making so little part techcrunch mentioned.