Camio Wins 24,000 Camera AI VSaaS Customer

By: John Honovich, Published on Jun 29, 2018

Silicon Valley Artificial Intelligent VSaaS startup Camio has announced it has won a 24,000 camera deal, a sizeable system for a small company.

Camio is one of the most technologically sophisticated video surveillance startups (see our testing Camio Natural Language Processing Tested and Ex-Googler Startup Camio Cloud Video Tested). However, the engineering-focused company has struggled at sales and marketing.

In this note, we examine the 24,000 camera deal, based on speaking with Camio, as well as examine the company's challenges vs competitors like Eagle Eye, Hikvision and Dahua.

******* ****** ********** *********** VSaaS ************ *** *********** *** *** * 24,000 ****** ****, * sizeable ****** *** * small *******.

***** ** *** ** the **** *************** ************* video ************ ******** (*** our ******* ***** ******* ******** ********** Tested *** **-******* ******* ***** ***** Video ******). *******, *** ***********-******* company *** ********* ** sales *** *********.

** **** ****, ** examine *** **,*** ****** deal, ***** ** ******** with *****, ** **** as ******* *** *******'* challenges ** *********** **** Eagle ***, ********* *** Dahua.

[***************]

24,000 ****** **** ********

***** **** *** **** ****** company, ******** ***********, **** be ********* ***** ** **,*** cameras ****** *** ********* starting **** ******. **** location **** ************ '*****'** ******* ***** *******, process ***** *** **** to *****'* ***** *******.

*** ******* **** **** won ****:

*** ***** ** ***********: 1) ***** **-**** *** packages *** ****** ***** times, ****** ******, ***.., and *) ****** *** of **-**** ***/***. *** main ******* *** ******** ***** **** *) ****-************* *** b) ************* ** **** analyses *** **** ******.

Revenue **** ****?

***** ***** ******** ** ******** revenue, we ******** **** ***** be ** *** ***** of $* ******* *** year, ******** $** *** camera *** *****, ****** 24,000 *******.

** ********, ***-**** ******* NVRs ***** ** ********* *** a ******** ** *** annual **** ** **** a *******. ** *** other ****, *** *** element ******* *** **** the ***** ********* *** Camio's ********* *** ******.

Camio **********

***** **** ******** ******** and ********* *** ******** in *** **** ****, Camio *** **** ***** this ****** **** *** some ****. *** *******, *** video ***** ******* **** *********** ***** ******** / labeling *** *** ** may ** ****:

Vs ***** ***

***** ***, **** * similar hybrid ***** ********, *** lacking *****'* ************* **** learning / ****** *********, has ****** *** **** significant ****** ********, ****** by ***** ***'* *** greater ********** **** *****, marketing *** ********** * more ********** ***** '***'.

Vs ***** *** *********

***** *** ********* **** spent *** **** **** aggressively ********* **** ******** and ********** ************, ****** their ******** ********** ******** with *********** ** **** to ** ******** ** many *******. ***** *** actually *********** *** ********* what ***** *** ********* would **** ** ** but ***** *** ********* ** market **** ********.

******, ****** ** ******* the ** ********** ***** block *** ****, ***** and ********* ***** ** prudent ** ****** *** out ***** ** ****** ***** technology.

Outlook *** *****

***** * **,*** ****** project *** ** ********** a ********* *** *** company *** *** ******* says **** *** ******** on **** ******** *** upcoming fundraising, ***** **** **** to ****** *** **** heavily ** ***** *** marketing ** * ****** that ** *** **** interested ** ***** ********. If **** ** ***, they **** ****** *** opportunity ** *****, ****** pocket *************, ***** *** opportunity.

Comments (33)

Sean Nelson

Nelly's Security
06/29/18 01:57pm

Congrats Camio. Incredible Analytics

Gerald Becker

06/29/18 03:25pm

Good for them! I love seeing startups making headway.

Undisclosed #1

In reply to Gerald Becker | 06/30/18 08:29pm

I love seeing any company make 'headway' - i.e. new and valuable features/platforms.

However, I do admit that I like seeing smaller, smarter companies that focus on deliverables produce stuff that larger companies (focused on market share) can't.

Fabian Muyawa

LONTECH SYSTEMS
IPVMU Certified | 06/30/18 07:09pm

This is a giant step forward in the market place.

Undisclosed #2

IPVMU Certified | 07/01/18 10:51pm

This blurb from the announcement:

The old proprietary, vendor-dependent systems common to the security industry could instead become flexible, on-demand, best-of-breed combinations of services, orchestrated around cloud data.

could say instead:

The old proprietary, vendor-dependent systems common to the security industry could instead become the new proprietary, vendor-dependent systems orchestrated around cloud data.

Undisclosed #2

IPVMU Certified | 07/02/18 02:56am

24,000 cameras/800 locations = 30 cameras per location.

That seems like a high number of cameras for a retail storefront w/o shoplifting or cash register concerns.

Anyone agree?

John Honovich

IPVM | In reply to Undisclosed #2 | 07/02/18 12:06pm

for a retail storefront w/o shoplifting or cash register concerns.

Camio says it was for a 'real estate company', not 'retail'.

Undisclosed #2

IPVMU Certified | In reply to John Honovich | 07/02/18 04:37pm

Camio says it was for a 'real estate company', not 'retail'.

Ah, my bad. I assumed that what Camio’s competitors were offering, namely

beefy on-site GPU packages for retail dwell times, people counts, etc..,

was something that was applicable in some way to what was required.

Any word on what the breakdown % is between inside/outside cameras?

Does Camio plan to do this install work themselves?

What’s the rollout timeframe?

Undisclosed #1

07/02/18 05:19pm

I wonder if IC Realtime is part of this win?

...and if not, does this annoy them?

Ella

Carter Maslan

Camio | In reply to Undisclosed #1 | 07/03/18 12:50am

No, this originated from Google. IC Realtime and Camio support each other as partners whenever there's opportunity for either of us. This was in-house with existing cameras.

Undisclosed #2

IPVMU Certified | 07/02/18 05:39pm

Is there really a 24,000 camera deal signed?

I don’t see such a claim in the announcement:

In planning video security for 800 properties, a leading real estate company asked a simple question: why aren’t better solutions available? Why is the state-of-the-art in physical security so far behind all the other software powering other business processes? That simple question kicked off a 6-month collaboration with the Camio team to demonstrate what’s possible with an open video processing pipeline built on modern technology platforms.

The key insight behind selecting Camio was that the SaaS approach to video surveillance manages operational costs to match both security risks and business needs.

Sounds like they were selected for a POC, unless of course the “6-month collaboration to demonstrate” is the whole system.

Of course, IPVM or Camio May have more information, but based on the release the scope of any actual deal is unclear, IMHO.

Carter Maslan

Camio | 07/02/18 10:14pm

The PoC ran from December through February. The 24,000 camera count is only upon full deployment among existing properties. Nothing in the agreement forces them to connect 24,000 cameras. That's one of the benefits of SaaS. Each month's bill is for that month's service.

Steve Lewis

In reply to Carter Maslan | 07/03/18 12:40am

Congrats Carter.

Undisclosed #1

In reply to Carter Maslan | 07/03/18 12:46am

"That's one of the benefits of SaaS. Each month's bill is for that month's service."

That is also one of the weaknesses of SaaS.... in comparison to TCO over the lifetime of traditional surveillance solutions.

SaaS changes the business model funding principle from Capex to Opex - therefore (imo) a listed 'benefit' like 'each month's bill is for that month's service' is not a direct benefit (as you are using it in comparison) but instead, a condition of the new business model only.

John Honovich

IPVM | In reply to Undisclosed #1 | 07/03/18 12:50am

That is also one of the weaknesses of SaaS.... in comparison to TCO over the lifetime of traditional surveillance solutions.

I am not sure your point here but the TCO of a 'traditional surveillance solution' is simply more front loaded, i.e. buy all equipment up front rather than 'renting' it.

It could be worse (i.e., it's a lot more money total), it could be better. One way it is clearly better for buyers, is that you can cut your losses quicker (i.e., stop paying if you don't like it). Yes/no?

Undisclosed #1

In reply to John Honovich | 07/03/18 01:08am

"I am not sure your point here but the TCO of a 'traditional surveillance solution' is simply more front loaded, i.e. buy all equipment up front rather than 'renting' it."

It is not 'simply more front-loaded'. This assumes that TCO is equal when there is a clear finite cost in a 'traditional' solution (disregarding maintenance costs as relatively equal in both biz models) while this is not always the case in a SaaS model.

TCO in a SaaS model is not a finite thing, but instead an 'ongoing' thing. Hence my argument that costs per month (Opex) vs finite costs (Capex) is not a 'benefit', but instead a difference based on business model. i.e. it can be seen as a weakness because the costs continue with use - potentially well beyond the TCO of traditional systems.

"One way it is clearly better for buyers, is that you can cut your losses quicker (i.e., stop paying if you don't like it). "

I agree that - in a direct comparison of business models - that this is a clear benefit.

Primarily because the lack of such a thing can't really be defended as something 'as good' (i.e. higher risk is always bad).

The fact that we label front-loaded (and finite) costs for systems as 'traditional' would suggest that some see benefits in this model.

John Honovich

IPVM | In reply to Undisclosed #1 | 07/03/18 01:11am

TCO in a SaaS model is not a finite thing, but instead an 'ongoing' thing.

TCO in a traditional model is not finite as well. Even though you buy the equipment up front, there are some maintenance costs ongoing.

The fact that we label front-loaded (and finite) costs for systems as 'traditional' would suggest that some see benefits in this model.

Of course, there are many benefits to the traditional model, that's why 99% of video systems still use it.

The bigger question is if and when the cloud model will overtake the traditional.

Undisclosed #2

IPVMU Certified | In reply to Carter Maslan | 07/03/18 03:20am

The 24,000 camera count is only upon full deployment among existing properties. Nothing in the agreement forces them to connect 24,000 cameras.

How many cameras/locations were in the POC?

How many are currently deployed?

Undisclosed #1

07/03/18 01:25am

"TCO in a traditional model is not finite as well. Even though you buy the equipment up front, there are some maintenance costs ongoing."

yes, I mentioned above that there are maintenance costs - and that they are probably close to equal in both models. Which allows us to ignore that factor in comparison examples of both biz models.

See my first comment again... my issue was with Carters claim that 'per month' billing for services used is a benefit. I'm not saying it can't be a benefit - specifically to those that lack the Capex required for large scale 'traditional' solutions... I'm simply saying that this very same thing could be seen as a weakness - based on fixed TCO vs not fixed.

From Carter's perspective, I imagine he does see this as a straight benefit.

I don't.

Had he mentioned that one of the benefits of Camio is that 'you can cut your losses quicker (i.e., stop paying if you don't like it)' I would have agreed with that and not questioned his choice of the use of the word 'benefit'.

Carter Maslan

Camio | In reply to Undisclosed #1 | 07/03/18 01:48am

Usage-based billing is a straight benefit when your service levels vary by time-of-day, scene activity, who's on duty, etc. It's not just a depreciated CapEx vs. OpEx accounting exercise when SaaS makes it practical to expand and contract your compute expenses throughout the day to reduce total costs.

Undisclosed #1

In reply to Carter Maslan | 07/03/18 02:21am

That sounds logical... and I am not saying that 'it's just a depreciated Capex vs Opex accounting exercise'.

However, 'expanding and contracting your compute expenses throughout the day to reduce total costs' should really be 'to reduce your incremental costs' - not total.

Can you describe an actual TCO for your product?

I don't think so - primarily because there is no end point to cost in your business model.

Can you offer the ability to expand and contract expenses based on usage (each month) and call that a benefit? absolutely - because the traditional biz model doesn't provide this option. But TCO will always be indeterminate in a SaaS model - simply because of the nature of the model.

Carter Maslan

Camio | In reply to Undisclosed #1 | 07/03/18 05:35am

There's no upfront purchase model that works forever without hard drive replacements, license upgrades, security patches, data migrations, access control admin, password resets, VPNs, on-site maintenance, integrations, etc... TCO calculations aren't simple regardless of whether it's SaaS or upfront purchasing. It's a false choice anyway. Both models are legit. Even this "SaaS" deployment has upfront costs, because the on-premise hardware reduces costs and improves quality.

Undisclosed #1

In reply to Carter Maslan | 07/06/18 04:08am

"It's a false choice anyway. Both models are legit."

Who said that both business models were not legit?

However, as to the first part of your comment, TCO is only a false choice in a SaaS model because there is no finite cost in this model.

In the SaaS model, TCO can not be adequately determined - because of indeterminate costs inherent in this model.

Stating that monthly billing for services is a 'benefit' is cherry picking that ignores any other perspective but your own.

Birger Kollstrand

07/04/18 09:28am

In this kind of deployment are the customers required to have redundant internet connections for secure full operation as parts of the functionality is in the cloud?

Carter Maslan

Camio | In reply to Birger Kollstrand | 07/04/18 03:25pm

I don't know their network topology but will ask them whether any redundancies are included in each property's connection. If the uplink connection were lost temporarily, then the video is still recorded on-premise and queued for upload upon reconnection.

Undisclosed #2

IPVMU Certified | 07/04/18 09:41am

Carter, can you explain what exactly were the trade-offs that led to the 2/3 cost savings for continuous AI vs scheduled AI?

Were these scheduled analytics for non-security functions?

After running a two-month trial across three properties, the first usage-based cost estimates for continuous AI video monitoring were 3x the targeted budget! That sobering budget discrepancy triggered a deeper look at the use cases that delivered the most business value. In most cases, the higher security risks corresponded to lower-activity periods of the day. So by scheduling the advanced analysis and indexing for each individual camera, the overall operational costs dropped radically (by ⅔). The on-premise video gateways still buffered 24x7 video to be indexed upon request and pushed the latest salient video events to the SOC video wall.

Carter Maslan

Camio | 07/04/18 03:39pm

Yes, ping me via email to carter@camio.com.

The trade-offs are specific to the use cases of each scene. But the primary approach is to defer work during specific periods of the day until it's either needed or requested while using the on-premise gateway as a "canary" that can alert even without the cloud AI. The secondary approach is statistical sampling.

Undisclosed #2

IPVMU Certified | In reply to Carter Maslan | 07/04/18 05:02pm

But the primary approach is to defer work during specific periods of the day...

Is the video still going to the (non-AI) cloud in real-time?

Is the additional cost for real-time analytics a pass-thru cost based on ACS (or other...) third-party processing?

Carter Maslan

Camio | In reply to Undisclosed #2 | 07/04/18 10:14pm

The on-premise AI runs continually while segmenting into 12s video events, so there's ~20s latency before reaching the cloud. That latency is a cost reduction technique that can be tuned up/down.

I don't know what ACS is, but Camio uses hooks for any optional 3rd party analytics and those hooks can be billed directly to the 3rd party or via Camio. But the Camio AI that ranks and filters events radically reduces the costs of best-of-breed 3rd party services in many cases.

For example, we're currently evaluating face recognition from deepvisionai.com and Camio radically reduce the costs by detecting humans and detecting faces prior to recognizing faces.

Undisclosed #2

IPVMU Certified | In reply to Carter Maslan | 07/05/18 02:32am

I don't know what ACS...

Meant AWS/ECS from Amazon.

For example, we're currently evaluating face recognition from deepvisionai.com and Camio radically reduce the costs by detecting humans and detecting faces prior to recognizing faces.

Since you’re in the pipeline, you might as well train your own ai/face recognition algorithm based on the third-party responses, and then take over completely when you are performing as well as them ;)

Carter Maslan

Camio | In reply to Undisclosed #2 | 07/05/18 07:22pm

That's the old thinking that creates the mediocre silos (zero-sum scarcity in value chain vs. abundance in customer needs). The security industry is looking more like the tech industry now. Good tech partners know how to collaborate for win/win bigger pies (even as frenemies). It's Bill Joy's "no matter who you are, most of the smartest people work for someone else."

Undisclosed #2

IPVMU Certified | In reply to Carter Maslan | 07/05/18 08:11pm

That's the old thinking that creates the mediocre silos...

What’s the new thinking, write press releases that lead one to believe that you just closed a 3 million dollar, 24,000 camera deal, from an undisclosed customer when the reality appears to fall far short of that?

Undisclosed #1

In reply to Undisclosed #2 | 07/05/18 08:29pm