Ask HN: Please grill us on our YC idea

https://news.ycombinator.com/item?id=3904673

by cjr • 14 years ago

9 25 14 years ago

news.ycombinator.com

Hey, we're at Haus in SF right now and wondered if anyone out there would help us out and want to grill us on our YC idea adifyit.com, our interview is Monday.

GoofyGewber• 14 years ago

That's a really cool idea, good luck!

cjr8• 14 years ago

Thanks!

Mz2• 14 years ago

If you haven't yet, you could try answering some of the questions here: http://news.ycombinator.com/item?id=3896121

Maybe answering them publically would get some feedback?

Best of luck.

semisight• 14 years ago

This is a very neat idea. A couple of questions come to mind:

Do you rely on an outside source for picture identification? If so, what's your plan if/when it fails? If not, do you plan to build your own in-house system?

Have you tested whether your code has any impact in page load speed (Y-Slow and company)?

Related to Q1, if you have your own in-house system, will you have your own database of pictures for feature extraction/matching, and how do you plan to make it large enough (and grow it fast enough) to accomodate the wide variety of pictures that people take and post?

I think that if you implement this well, you potentially have a real winner here. Best of luck to you!

cjr8• 14 years ago

Thanks for your reply. We are building the picture identification in-house, we have a crude working demo as a proof of concept that the tech is good enough.

We've used the Amazon product API to build our databsae, there are ~80m products in there. We're starting one vertical at a time (the algo works best for highly textured products such as watches/book/cd covers so we've started there). We build an in-memory index of image-descriptors which is queried when trying to match a product. As we start seeing other product type, we'll expand beyond amazon.

As for scaling, the index is the one component that isn't trivial to parallelise. Once we get to a stage where a single box can't store it in memory, the first step would be to shard it, with different indexes per product category. Query images will be run through a product-categorisation algo to determine which indexes they need to be run against. This is how we think we'll approach this, from previous experience, but will investigate properly (if) we get large enough.

In terms of affecting page load, our plugin will load async, as will the results returned from the image matching algorithm, so it should be fairly negligible impact.

biopharma_guy• 14 years ago

Go to this thread below and grill yourself with these 34 questions. http://news.ycombinator.com/item?id=3896121

You perhaps will find yourself where you are in the application. By the way, the idea is cool. But does your algorithms advanced enough to distinguish between items that can be bought (commercials) compared to those can't (human face, nature photographs etc to example a few)?

cjr8• 14 years ago

Thanks for your reply. Yes, the algorithm would match against product images, so most likely faces and nature photographs wouldn't match any images in our product database.

Here is an example of what the ads would look like on our site dump.ly:

http://dump.ly/y3uyGE2hpn2j

debacle• 14 years ago

There's no link to buy Scarlett Johansson.

tonyjwang• 14 years ago

Under "Get started in minutes" Its should be It's.

Cool idea.

cjr8• 14 years ago

Cool thanks, fixed.

Mz2• 14 years ago

Oh, and clickable: http://adifyit.com/

hardik988• 14 years ago

Great concept - could be the future in advertising.

But like any ad network, how do you guys plan on keeping up with splogs (spam blogs) abusing your service ?

dmmalam2• 14 years ago

As our ads are affiliate links (CPA), it shouldn't be a problem as a user has to actually buy something for payments to go through.

tnorthcutt• 14 years ago

Adify.it links are only shown when users engage and mouse-over a product.

What about mobile?

cjr8• 14 years ago

For mobile as there's no concept of hover, we'd put an always-on button in a user-specified corner of the image. When the button is tapped our ad overlay would show.

damoncali2• 14 years ago

The "Why trust us..." headline made me wonder if it works very well. Why wouldn't I trust you?

cjr8• 14 years ago

Thanks for your feedback. There is an element of trust involved as we are going to modify the look/feel of the host site. But if this causes potential customers to wonder if it works well, then we should probably reword it.

damoncali2• 14 years ago

For what it's worth, here was my reaction: At first I was thinking "cool". then I read that headline, and thought, "I hadn't thought of this as something I need to trust people with. In fact, now that I think about it, I bet that's really hard to pull off." Then I thought about the crazy non-contextual intellisense stuff out there and was wondering if the whole mess is even possible without turning the website into a spamfest.

teem• 14 years ago

http://stippleit.com/

bkyan2• 14 years ago

How do you determine where the ad dot appears in the picture?

dmmalam2• 14 years ago

Our algos can work out where the product is in the image. Technically it's the centre of the convex hull of the set of matched keypoints

igorsyl• 14 years ago

Close but no cigar http://taggify.net

bkyan2• 14 years ago

How did your interview go?

Ask HN: Please grill us on our YC idea

Related Stories