[mfrye0]

If you do this sort of thing often, I'd love to chat further. I'm basically trying to automate this sort of manual research around companies with a library of deep research APIs.

Had a show HN last week that seemed to go under the radar: https://news.ycombinator.com/item?id=45671087

We launched corporate hierarchy research and working on UBO now. From the corporate hierarchy standpoint, it looks like the Delaware entity fully owns the Estonian entity. Auto generated mermaid diagram from the deep research:

  graph TD
    e1[BuildJet, Inc.]-->|100%, 2022-12-16|e2[Buildjet OÜ]

[keyle]

It's fairly easy to know how to poke around these businesses. Look up the people, the business, and the product. It's less fun when it involves linkedin. Every country has a database of business numbers to name and rough documentation. Dns look up can reveal some information. Social media typically finish the rest. These "founders" are often serial founders, with a ton of abandoned projects and a trail on product hunt, and other websites.

In this case, what really gets to me is the basic template website they're using; with image carousel but only one image... and the fact that they appeared to have paid influencers on youtube to shill their product.

It feels rushed, and not in a good way.

[mfrye0]

I noticed the template too. Someone mentioned recently that's actually a good risk signal - scammers often use the same site structure across domains.

On the research, you're absolutely right. It fits that sweet spot where it's just easy / boring / tedious enough to automate with the current generation of LLMs.

[davsti4]

So, I should expect to see a new product launch of DoughnutKVM, to "round out your infrastructure woes", complete with vibe coded app interface and AI generated product images, here in the next week? ;P

[neilv]

Looks like you have a potentially great business for corporate compliance, if you can answer with plausibly high confidence (or indemnify?).

I only occasionally research companies, and it's from an engineering&product perspective, aside from corporate ownership compliance. (For example, I was asked to vet a little-known company as a prospective partner, for building our cloud infrastructure atop theirs. One of the first rapid low-cost, high-value things I could do, besides looking at their docs and trying their demos, was to skim through the history of business news about them.)

[mfrye0]

Interesting. That's actually where we started. We were doing automated research on vendors from a TPRM perspective and looking for data points around organizational security / reputation. Examples - if the company had been hacked before / how they responded, do they have a CISO, nth party vendors, are they SOC2 / FedRAMP certified, etc. Basically, predictors of risk / stability.

We realized the underlying business graph was the bottleneck though, so that's been our focus for some time. With that in place, we're now coming full circle on the risk research standpoint.

On your comment about confidence / liability, we're actually having conversations around that now and getting feedback. First step is exposing all the research and evidence directly to build trust, which is what we're doing now for the new corporate hierarchy system.

[travem]

When I used to work in credit control and accounts receivable, the use of D-U-N-S numbers was how we tied a lot of this information together. It is similar to how SSN are used by credit rating agencies but for businesses and global (unlike the EIN).

[embedding-shape]

If you want to feature a governance structure infamously hard to get right and impressive to use as an demo, IKEA/Ingka would be an good example.

[mfrye0]

Good idea! I picked a random California Ikea entity (IKEA US RETAIL LLC) and ran it through the system. Here's the output - current goal is to get to ultimate parent.

## Summary IKEA US RETAIL LLC is a limited liability company. It is wholly owned by IKEA Holding U.S., Inc., and ultimately controlled by Stichting INGKA Foundation, a Dutch foundation that owns Ingka Group.

## Graph

  graph TD
    e2[IKEA Property, Inc.]-->e1[IKEA US RETAIL LLC]
    e3[IKEA Holding U.S., Inc.]-->e1[IKEA US RETAIL LLC]
    e4[Ingka Holding B.V.]-->e3[IKEA Holding U.S., Inc.]
    e4[Ingka Holding B.V.]-->e4[Ingka Holding B.V.]
    e5[Stichting INGKA Foundation]-->|100%, 1982|e4[Ingka Holding B.V.]

This is the permalink to the deep research result: https://savvyiq.ai/playground/entity-hierarchy/siq_31ro4EDce...

[jychang]

Login gated

[mfrye0]

Sorry, habit. I've been debating on exposing these publicly, but they're expensive to create. We have a public interactive demo here for now: https://savvyiq.ai/products/entity-hierarchy

Here's the live mermaid editor version for the Ikea example: https://mermaid.live/edit#pako:eNqNkV9PwjAUxb9KcxPfRrO17E_3Y...