Hacker News new | ask | show | jobs
Show HN: I made a cloud security product that actually saves time (argos-security.io)
63 points by david_obrien 1570 days ago
I have been a Cloud Security Consultant for over 13 years and all that experience I put into building ARGOS Cloud Security.

Probably the most common thing I've seen was teams spending loads of time manually investigating alerts from security products. Detecting security issues is not what it’s ever been about, but about the investigation of these detected issues. If you don’t have time to investigate everything, then what’s the point?

Organizations spend time and money on security consultants to help them investigate their environments, but these are often only point-in-time engagements, where, once done, nothing "reusable" is left behind. The cloud environment might be in slightly better shape, but maybe only for a short period, before the next contractor must come in to start the process all over again.

I, and many others I've worked with, always believed that the most difficult part of cloud security is distinguishing security issues from "just best practice violations". Using an "everything must be compliant" approach never really works and usually means things just don't get done. If everything is red, what do you do?

So, people spend immense amounts of time trying to figure out what cloud misconfigurations are actual security issues, what else is put at risk because of something that is misconfigured, and how it can be fixed. All that information is important to eventually prioritize and fix issues.

ARGOS investigates the things it finds, that's what it does. It's not about "finding more", it's about understanding what misconfiguration is exposing a cloud system to the internet and what else in your cloud environment is that misconfiguration indirectly putting at risk.

One awesome side-effect of this is that we even draw something like an architecture diagram of the environment "around the misconfigured resource". This specifically is something I know most past customers of mine lacked. Architecture diagrams were hard to come by, and never up to date.

It's easy to try out if you have access to a cloud environment like Azure or AWS, GCP is in a very early version right now. 20 minutes or less even is all it takes to get everything going, including a one-click Slack integration. First results should be in your dashboard minutes later and are continuously and automatically updated.

Really looking forward to people's feedback. Thanks all!

14 comments

I don’t want to be that guy, but when I access you page I get warnings about the site not being secure. On mobile I’m not able to see why, but on my desktop I can see that some of your images are being loaded from http: which is causing alerts about insecure content in Firefox.

I’m only bringing it up because when I’m looking at security tools, I don’t want to see security issues (even if it’s just their marketing website)

Fair call. I will definitely need to check that. It seems like Chrome and Edge don't show those warnings. I will install Firefox and check what's up. Thanks.
If you reload the page a few times it goes away?

I've been noting this with ff lately (at least on mobile). The sites will be fine in chrome but have the error in ff. Unfortunately android ff doesn't let you see details about the error

If you reload the page the issue does go away... but only technically.

The issue is the page references a few images over http, which redirect to their https form with a permo redirect so the browser knows the new location when you refresh. Your browser also now has the image cached too so doesn't need to request it

One of those two seem to make it stop flagging the problem. I guess the browser only complains if it needs to make the request

@OP: I'd do a search and replace http://domain to https://domain and try to use relative urls where you can in future (I am aware this was probably a WordPress thing, not a you thing!) - My go-to for this is https://wordpress.org/plugins/better-search-replace/ for the content side and a good ol text editor in your theme files if needed, should be good to go :)

Thanks, I'll definitely look into that.
Just a word of warning that Argos is already massive established retail brand in the UK (and some over sees territories). I appreciate naming things is hard so don’t take this as a judgement about what you should do, just be aware that you might run into some trouble there.
I'm British, so I'm aware of the retailer. Thanks for the warning, we've been assured by many legal professionals that this should not be an issue. We tell everybody that we are not associated with them and that we are actually named after Argos Panoptes (https://en.wikipedia.org/wiki/Argus_Panoptes), the all-seeing giant.
I guess I'd caution that there might not be an issue now in a strictly legal sense, but come trademark registration time Argos-the-retailer might decide to put up a fight.

My more cynical side wishes to express that this would make the lawyers very happy.

Trademarks are registered within a specific category of business. That's why Apple, the music label, could exist next to Apple, the computer company, all those years back. It's also why Apple got into a trademark dispute all those years down the line when they got started in the music business.

The http://argos.com/ people don't have trouble with the ARGOS trademark (there was a lawsuit about this!). OP's product might not be so lucky, though.

If OP can get a trademark registration for the right category through, they should be fine using the name Argos from what I can tell. Either Argos or ARGOS might object to that trademark and make that very difficult, though.

That’s not quite how it works. Companies argue that there’s no brand confusion in order to coexist with similar trading names and use market sector as evidence but there isn’t any specific law that says this has to be the case.

To use your Beetles vs Apple Computers case study, Apple Computers actually settled with the record label and part of the agreement of that settlement was that Apple wouldn’t enter the music industry. It wasn’t a case that the record label lost, it was a settlement. Hence the follow up case between the two after Apple released iTunes (or whatever it was that triggered their “arrival” in the music industry).

Panoptes would not be a bad name, even though, as you would imagine, it is already taken by a few different organisations.
There’s also Argo CD in the cloud world
I don’t know what to make of this. It seems like this is an enterprise product meant for companies with procurement and legal teams. Why I say that is because there is a pricing link in the top menu (but surprisingly not in the footer links), but the pricing page shows nothing about the price at all. I have no anchor and no inclination to check out the rest of the website because of this. Either expose the pricing upfront or just put one single “Contact us for pricing” button. There is no need to put two blocks for pricing because there’s no way anyone can figure out what these mean on the costs for them (since, for all we know, each customer could be given largely different price quotes based on their name, size and other factors).
Thanks for your comment. As I mentioned to others, I followed what other companies are doing and what actual customers (not just Enterprises) asked me to do.

If I just put a "contact us for pricing" button onto the website, would that make you check the rest out and potentially sign up for a trial?

Also, ARGOS is still early and the pricing, to be completely transparent, is still very custom depending on the value a company communicates to me. It's not meant as a trick, otherwise I wouldn't give out a free trial for 14 days to everybody without even asking for a credit card. It's genuinely a "I don't really know yet how much to charge as a baseline".

You don't understand where you are in the business lifecycle.

You're very early. Paying customers are the most important thing for you to get.

However, it doesn't matter how much the first 1% (of your target market) pay, now or in the future. (If you're successful, they're only 1%. If you're not successful....)

Let's assume that success is 100k-1M paying customers.

Anything that gets in the way of someone being one of your first 1,000 paying customers is bad. Not telling the price upfront is one such thing. Trying to charge "market/what it's worth" is another.

Now is not the time to optimize pricing. In fact, you should tell the first 1-2% of paying customers that they will always and without any effort on their part get better pricing than anyone else going forward.

Yes, you should ask them what they think other people should/would pay....

I appreciate that comment, thanks. It does feel a bit like a catch22, that I am definitely trying to solve.

ARGOS does have early paying customers, and of course they are heavily discounted (and they know).

You mention not telling a price is an issue (and so have others, point taken), but also say I shouldn't focus on price right now. I'm trying not to. So, what's your recommendation? Should I remove the page? Rename it to something like "Plans" and not "Pricing"? But then I'm still not telling you how much it will be. I know how to deal with this when talking to someone, but what I'd be really interested in what you would expect on a website, knowing what you told me.

Seriously, thanks!

Put low prices on the pricing page together with a note along the lines of "prices/terms are negotiable". That page should probably also mention that customers now will always pay less than later customers.

Be as friction-free as possible. When you talk with someone, ask if there is anything that would have helped. (The problem being that you want to hear from the folks who didn't talk with you.)

Note - while you want an interesting number of paying customers, you don't want any bad customers, where bad means "suck up a lot of time, provide no benefit."

Be flexible up to the point where you feel that you're being take-advantage of.

For example, if you're offering 15 days free and you're asked for 30, that's probably okay (and may be useful information telling you what you should offer). However, if someone demands 6-12 months, thank them and tell them to go elsewhere, nicely.

Thanks, those are great tips and I've just implemented some of those.

Thanks again, this is super helpful!

The landing page looks nice. I feel like, based on your comments, you really are trying to build a product for customers - and not building towards some esoteric sales figure. Some comments from someone who's worked for a number of "enterprise" security vendors over the last decade...

Your product is competing in the CNAPP (Cloud Native Application Protection Platform) space. I'm sure you already knew this, but you may want to look at how companies have merged CSPM/CIEM/IaC/etc tooling to get to this "Gartnerized" name and just know who's who in the competitive landscape.

14 days isn't long enough to make a decision for most organizations unless they are really small. I'd suggest leaving that in place but also extending that to 30-45 days (1 month workflow) for prospects that will, in turn, go through additional resistance to validate they're truly viable customer candidates. Those things could be a 15 minute qualification call where you have a list of things you want to review with them to understand the fit (good feedback loop for you in features).

Standardize on your pricing - you need to understand this so you're both not losing money and so that you don't turn customers away because of the unknown. You can always have a custom pricing tier or an "Ask Us" option. If you don't know what your baseline cost for a customer is - then figure it out quickly.

Demos can go a long way - since you're a small shop you need something that won't soak your time, so I'd suggest taking your customer's top couple pain points and recording a nice looking demo that won't cost them more than 10 minutes to watch. People in this space want to see the product. There are so many security tools that a lot of buyers already have too many - so they want to see what it is and many customers are also looking for ways that new tooling can be integrated into existing workflows. There is no "single pane of glass" anymore for customers and instead it's morphed into, as I like to refer to it: "single pain of glass", because every vendor claims you only need them. Right.

If you have any questions I'd be happy to chat - [my_hn_name] at counterbrea dot ch. Best of luck!

I am indeed. I've seen those issues one too many times as a consultant, and I never got to really(!) fix them as I was only ever brought in as a band aid really, not strategically.

Yeah, CNAPP is the new acronym for where we probably best fit right now. I will continue checking out some of the other players.

I initially had a 14 day trial, then people (not potential customers) told me that would not be long enough, so I increased it to 30. That did not make a difference at all. Smaller companies still signed up and converted, larger companies "ignored" any time limits and just assume that those don't apply to them.

Based on all the comments now I wonder if I should have a "base" plan that has the cost on the website and an Enterprise / Custom plan with a "Contact us" button. I pretty much know what a customer costs "on average", so that's not difficult.

Curious, did you see the short demo video on the website? There is a short one on the main site "below the fold" and there are two more in the "Resources/Videos" section. I am working on a longer version that I can send out to potential customers instead of a demo initially. Yeah, I've seen and felt the "single pain of glass" many times :D

I'd love to take you up on that offer, thanks, a lot!

"Price on Application (also available after subscribed to the Trial for self-service customers)"

Please dont do this... I even clicked on the follow through link, then you want me to sso... still no price. Save us all some time, it is in short supply.

It's a fair comment, and I used to have prices on the website and then actual customers, while they were on their way to becoming customers, told us to take the prices offline. Their procurement teams really didn't like prices on the website. Doomed if you do, doomed if you don't. The trial is completely free, we don't ask for a credit card, only when you decide to want to become a customer, you can actually see the self service prices in your ARGOS dashboard. Those prices are also available via the Azure/AWS marketplaces.
If your entire target market is enterprise then I suppose this route works well but if you want to hit smaller organisations it deviates a lot from what we're used to. GitHub, Slack, GSuite, they all provide prices and are successfully sold to small organisations and enterprises.

As a non-enterprise, there's no point in me signing up for a trial of a tool that may be orders of magnitude outside of my budget. A lot of my decision is based on the value proposition and I can't understand the value I might get if I don't know the price. If it's $10 a month and does what it says then I probably won't think about it, if it's $1,000 a month then that's quite a different decision process. I don't want to bother with a free tiral if I can't even see the price. Am I just booking to test drive a car I'll never be able to afford?

Again, fair comment. Frankly, I'm still figuring out what this would be worth to smaller organisations. ARGOS really scales by AWS Account / Azure Subscription / GCP Project right now and I'm currently charging "for each...".

I know what Enterprises are paying for ARGOS and some larger SMB, but not clear on what value smaller orgs or even startups assign to this problem.

keep in mind, AWS best practices (and IAM limitations, for those without resources/time to finely craft the boundaries) encourage account sprawl...

Not all customers spread out like we do... but I manage over 50 AWS accounts, and our DevOps team is 7 FTE... Our application/situation is admittedly unusual, but I could easily see a small business using 10 accounts to manage their org and a single "product"

I used to be a consultant building exactly those designs, limit the blast radius, have separate accounts etc.

"In the earlier days" of ARGOS that's exactly why I didn't charge per Account, but different ways (tried # of resources, then % of spend) and people were always confused.

Similar to the "take the price off the website" that customers told me, me charging per Account is also what customers asked me to do.

What do you think would a good unit be for a product like this? I'm happy to try anything really, as long as it helps companies be more secure.

That's completely reasonable, thanks for the insight!
There's a ton of companies doing CSPM products these days, and now it looks like the trick they're trying to achieve is integrating their detections into their ecosystem (notifications, automated remediations, custom policies, infrastructure-as-code scans, endpoint agent integration etc). Wondering what your roadmap is and how you'll take these big guys on?
Great question. First, I feel most of the integrations (aside from Slack) are really asked for by Enterprises. Smaller orgs don't really ask for those. Roadmap is absolutely around context, even more context. We want to get to the point where we're almost a cloud security lake (can't come up with a better word) where we know about everything that happens in the environment and allow you to make decisions almost on the spot based on what we show you. (Lateral movement possibilities by an attacker, what is actually running on those workloads and in what versions, who/what has access to them, etc) We're also looking to make our remediation engine more intelligent (not AI), but potentially allow you to give us a script we should execute to remediate, or trigger some other SOAR engine even. Right now remediation is whatever we think is a good remediation (and rollback) for an issue, but that's not always what you might want. Does that give you a view into my mind for a roadmap?
As an IDS/IPS software architect with a mind to use my private cloud for a checkout on security, I did the following on iPhone FIREFOX.

1. read the main page (noted accessibility issue in other comment)

2. Read the footer

3. Chooses and read FAQ

4. Choose and read Features

Almost went to Feature as first click. Probably wish I did.

It did feel that it was targeted toward enterprise and smartly so. I am inclined to contact but it’s a private cloud with propensity for walled garden (nearly airgapped).

If I were a real enterprise with private clouds, I too be curious about this functional set and whether they can be applied toward the airgapped cloud.

Nicely and tersely written for the target SEIM, CCIT and cloud-owner audience.

No bite here as my private cloud isn’t up for peerage, external, that is. And my customers have private clouds too.

Again, very nice from this angle.

Thanks for that detailed response. ARGOS definitely feels more like an Enterprise / larger SMB product, unfortunately, I think this is mostly due to where people are really feeling "the need" for cloud security. I had many conversations with smaller organisations / startups and most of them said that they could not spend money on something like cloud security (AV and Firewall was #1 spend in those conversations, very interesting) or sometimes even didn't believe that cloud security was a problem ("doesn't Microsoft/Amazon take care of this for me?").

So, as much as I'd like to help smaller organisations, it seems only at a certain size do people really recognise this as a problem.

What do you think? Also, what made the site "targeted at Enterprise" in your opinion?

Also, yes, ARGOS is, for now, only public cloud, no private cloud.

Thanks again.

What made me think it was for enterprise was as the few other commentators have said: pricing, wordings, and …

lack of the superficial First wave demo. This is the thorny issue that nearly plagues every marketing/sales.

Assuming that my own restrictive criteria were cast aside, I would be even more intrigue by some example eyeball-catching (and perhaps semi-interactive) faux outputs. For me, this would prompt me to contact you for more details of which i expect a set of authentication info for me to visit and look around as well as take a call for some Q and As.

Again, this last step is the fishing lure that enables those who are severely time-constraint to make that “last judgement” to bite that marketers and sales ever so want to cast … and snag.

However, enterprise folks may have more time on their hand (as opposed to SMB-like folks) and may make that push to elicit for a more personal feedback … without that semi-interactive demo.

Of course, the real danger to any demo is the lack of differentiation from what they already may have could translate to lost sales opportunity.

Thanks for the further details.

Did you see the short video on the first page? It's a bit further down and might need to be pulled up. We have two other videos under "Resources/Videos" as well.

I am currently working on a "longer demo video" that shows ARGOS's full potential and differentiation.

During my initial and only viewing of the landing page (save for FAQ and Features), no, I did not see a video.

Will visit again.

I'd love to see additional pricing models for security consultants and MSSPs. This is a fantastic idea and I am curious to test it out! I think the same model of doing SOARlike work for alerts would also be effective in areas like MS's security portal(s) utilizing the graph security api. I have been using Threat Dragon to map network risk boundaries manually. Integration with Rumble might be a good future thought to do similar automation for non-cloud environments.
Thanks for that feedback. We have MSSPs as partners already and the pricing there is pretty much "custom". I haven't thought about "individual security consultants" yet. I assume they would rather prefer a "point in time" scan to what we currently do (continuous scan).

I did actually think about reaching out to Rumble at some point. They seem to have figured out the whole "unauthenticated workload info" that I'd love to see within ARGOS.

Corner case: iPhone/iOS accessibility setting for reading in a bit larger font: your first line header of main page got truncated on left and right side of portrait view.

first glance issue.

Something that a paid “online cross-browser test” can quickly exposed.

Please disable automatic capitalisation on the "Request Your 30-minute ARGOS Demo" email field.
oh wow, never noticed. I will see if I can disable that. That's an odd one.
Looks great! Congrats on building it! How long did it take ? And how big of a team ?
Thanks. Appreciate the comment. Overall it's been in dev for the last 18 months with some paying, early customers across Australia and Europe for the last 12 months. It's a small team, really only 3 people (plus an intern). We want to be as lean as possible, not too much overhead with roles that aren't really required. As you've seen, the product is built to be mostly self-service by customers, no need to talk to us.
Great accomplishment! Good luck !
Looks like the server is down.
Apologies, that should be fixed now. Bit more traffic than anticipated.
I had a moment there, I'm good friends with another David Obrien, and thought my worlds had collided.
why in $DEITYs name is there an annoying multisecond fade out when clicking to another page, followed by a long delay to render the page. i was stopped right away from looking further.
Probably too much reliance on web designers. ;) I'll take that feedback and make some changes.
Yea, I’m general the design is a bit over the place. Large text on mobile, annoying chat bot, etc
I had to "save a bit" on the web designer, sorry. Focus was more on the product. The chat bot should have already been removed. I will do that asap.
That bot is gone now. Thanks for letting me know.
Interesting tool...how does this compare to other CSPM tools?
Thanks. CSPM products usually are great at finding any misconfiguration (compliance violation) in a cloud environment and mapping it to a Compliance framework. This is important, compliance I mean, but I've seen numerous organizations struggle to find actual security issues within all those "compliance violations". Don't get me wrong, configuring AWS Account contact details, resource tags and not underutilising resources is important, but not really a security issue.

ARGOS takes (currently) a subset of controls from well-known frameworks (like CIS, NIST, etc) and maps them to security frameworks like MITRE Att&Ck, but also automatically investigates each misconfiguration to see if there are other things in the environment exposed because of this one misconfiguration (i.e. "is an AWS RDS exposed because its SG allows access from an EC2 that is publicly exposed?") and then draws a diagram of the immediate environment around this misconfiguration, literally showing a "path" (cyber kill chain) that someone could walk if they were to make it into the environment.

ARGOS doesn't claim to find MORE things in an environment, we save you time where you'd have to manually investigate each alert, we show more context than just "here's a single misconfigured property" and we even have remediation baked into the platform.

Really, ARGOS probably doesn't 100% fit into the CSPM category. We have users that use a CSPM AND ARGOS.

Most tools attempt to a mix of both and at least claim that they could detect security violations (especially since you can customize which controls you want to look for).

Your site could really benefit from explaining what your differentiation is. Drawing the "path" sounds very cool, especially if you can do things like showing how an exposed VM could compromise an S3 bucket (via the VM's role).

Thanks. Yeah, our website needs to be clearer on the differentiator. Part of this Show HN was to also see if people "get" what ARGOS does.

Right now the path only takes network into account (no, we do not look at flow logs). Lateral movement via IAM roles is on the plan for next quarter.

Looks like a cool product.

How does this compare to Checkpoint CloudGuard?

Thanks!

Based on my understanding of CloudGuard we differentiate again by the "automated investigation" we do. It's not just a "you have misconfigured this thing here" but then also "because of that misconfiguration this following resource is now exposed to the internet (we checked, it is!) and it's exposing these other resources as well, look, here's a diagram that shows you this."

This is what products like CloudGuard expect you to manually do. They approach security from a compliance point of view. "You are violating control XYZ from framework ABC, so you are less secure." Unfortunately, it's not that simple.

An AWS EC2 instance should not have a Security Group on it that allows RDP from the internet, that's true and should probably be flagged, but in itself is not a security issue. It only is if the EC2 has a public IP (or a public Load Balancer), the VPC has an internet gateway and there's no NACL blocking the traffic. This is just one part of what ARGOS checks, what others just don't and expect you to do manually, for each of their hundreds and thousands of critical detections.

Thanks David,

> Based on my understanding of CloudGuard we differentiate again by the "automated investigation" we do. It's not just a "you have misconfigured this thing here" but then also "because of that misconfiguration this following resource is now exposed to the internet

CloudGuard does this exact same thing. WE just had our AWS env. scanned and they did tell us that one of our S3 buckets had a wildcarded principle. It documented the exact bucket name and the policy we used.

I agree that the diagram is something I haven't seen from other vendors in the space.

But I am still struggling to understand the differentiation of the other features.

A wildcarded principle by itself doesn't mean too much. There are settings outside of the S3 bucket access policy that can mean "it doesn't matter what the bucket thinks".

This here is something other products typically don't check, and because of that create a lot of noise that a person has to check through. https://www.youtube.com/watch?v=kMi5PSyFu8s

Other products only look at properties in isolation. As I mentioned the SG rules only become security relevant if many other things in an environment are also true, ARGOS checks for those, others typically don't. Only one example of our "context awareness".

The diagram shows the "kill chain" of how someone could laterally move through your environment, again something others typically don't do.

I said it before, we don't find more, we help you find the ones in the noise that actually matter from a security point of view.