The role of OSINT in finding missing children with Micah Hoffman

Matthew Stibbe (00:01.442)
Welcome to From the Source, the Blackdot podcast. I'm your host, Matthew Stibbe, and today I'm delighted to welcome Micah Hoffman, who is, and I've got a long title to read here, founder and co-owner at My OSINT Training and director of capabilities and innovation at The National Child Protection Task Force. Micah, great to have you on the show.

Micah Hoffman (00:21.812)
Well, thanks for having me on Matthew. It's good to be here.

Matthew Stibbe (00:24.45)
And how is the weather there in sunny Washington?

Micah Hoffman (00:27.9)
It is sunny and it is very cold and we've had more snow than we've had in probably nine to 10 years. So I'm ready for spring.

Matthew Stibbe (00:34.414)
Wow, really?

The last time I was in Washington it was freezing but I can picture that lovely crisp cold. So, well fantastic. So let's start off by asking you a question I ask all my guests. What are you geeking out about at the moment? What's on your mind?

Micah Hoffman (00:54.293)
So my current 'geeking out' is really doing research in open source intelligence. I love getting into the bits and bytes of websites and applications and looking for places where we can find interesting pieces of open source data that we can then make a tool to extract or to better visualise. And then my favourite thing is to make a tool that's releasable to everybody and

get that out into the wild so other people can do some really good things with it.

Matthew Stibbe (01:26.274)
And you do that how?

Micah Hoffman (01:28.69)
So I'm actually using AI a lot. Most of my work is in the web browser. And when I'm in the web browser, I'm using web development tools. It's built into every browser. And I use the web developer tools to look at the network connections, look at the calls that the website's making, look for hidden data inside of the web page. Then once I've identified that, I feed that over to my good AI buddy

and talk to them about how I want that extracted, visualised, and then ultimately packaged into what we call a bookmarklet. Then I take that, I put it on our website, I make a blog post, and everything's done.

Matthew Stibbe (02:10.863)
And this is on My OSINT Training, these bookmarklets?

Micah Hoffman (02:14.548)
That's correct. On tools.myosint.training, we have over 40 different bookmarklets for social media platforms, financial platforms, a bunch of general ones, and even ones that can help you do research into open source opportunities within certain webpages.

Matthew Stibbe (02:33.486)
As a sort of curious outsider, can you give me an example of a bookmarklet and what it does?

Micah Hoffman (02:40.68)
yeah. So we have several bookmarklets that when, for instance, let's say that you visit a TikTok profile. You visit the profile in your web browser as an authenticated or non-authenticated user. TikTok and many other websites will send more information to your system than what's rendered in the page. What our bookmarklets do is when you click it, it looks in the page source code and says, hey, look,

Matthew Stibbe (02:47.054)
Thank

Micah Hoffman (03:09.66)
on this TikTok page, this profile, there's a creation date and time. I'm going to extract that and show that to Micah. And so it will look in the page. It'll extract that. It'll extract the TikTok ID, which is that unique ID for that account and some other relevant data. And then it pops up a little box and says, hey, on this page, here's some extra data that's in your browser already that you might want to know for your OSINT investigation. So that's just one of them.

Matthew Stibbe (03:37.134)
So it's information that Information that the browser has already pushed down, but it isn't necessarily visible or obvious in the user interface?

Micah Hoffman (03:45.384)
That's right. Many of the bookmarklets take advantage of just extra data that's in your web browser already, not making any extra calls to the website. We call those passive bookmarklets. The other ones that we have are more active. And through some additional research, we found that if we make an extra call to LinkedIn or to another website, we can sometimes extract meaningful data that can be extra helpful. I was just working on one.

for the... I forget the platform. I was just working on one and hang on.

Micah Hoffman (04:25.44)
totally skipped my head. If you can just cut that out. So I was just working on one and a person came to me and said, Hey, you know, what country is this? it was TikTok. All right. There we go. So I was just working on  on a bookmarklet and I had a person from the community say, hey, listen, this other website can get me this information, but your bookmarklet doesn't. I did some additional research again, just using the free web developer tools and

Matthew Stibbe (04:36.622)
It they go.

Micah Hoffman (04:54.118)
I found some extra data in there, but in order to extract that data, the bookmarklet, through the user's browser has to make additional queries to the website. Because it makes additional queries, we call that an active bookmarklet. And on our tools.myosint.training website, we note, is it active? Is it passive? Because some people are not allowed to make those additional calls or are just wary of it, are concerned.

Matthew Stibbe (05:21.102)
Because of the source IP identification?

Micah Hoffman (05:24.334)
Exactly, or because, you know, their rules of engagement for what their company or their government or their law enforcement organisation says they can and cannot do is very particular. If they just browse to a page, that's fine. Once you start making requests on behalf of somebody's browser, it can make people a little bit skittish that, what else is he doing? But the coolest thing.... Exactly. And the coolest thing about some of these bookmarklets is that

Matthew Stibbe (05:46.936)
You need to know if it's going to do that. I get you. OK.

Micah Hoffman (05:53.184)
you have all of the code. So when you use one of these bookmarklets, you can take it to your favorite AI and say, tell me what this does. What is it going to do? Does it log any data? there any issues I should be aware of? And it will decode it because nothing is encrypted or nothing is closed source. It's it's all right there.

Matthew Stibbe (06:13.506)
Fantastic. this is something that is on My OSINT Training. Tell me a little bit about what My OSINT Training does.

Micah Hoffman (06:14.569)
Yeah.

Micah Hoffman (06:23.646)
Sure. So we're a online training academy. Most of what we do is myself and three other, well, two other authors and my partner. We make OSINT courses and we record them. We put them up there and people can order one course or all of our courses and watch them whenever they want to learn about open source intelligence, to learn about all the different aspects of it, whether it's...

using, whether it's setting up a system, protecting yourself, going on to Instagram, TikTok or some other platform. And we find that people all around the world are interested in learning how to do this. How to do open source intelligence.

Matthew Stibbe (07:07.118)
It's a fascinating thing because there are so many ways into the profession when I talk to people on this podcast, they've come into it from from librarianship or from whatever, whatever. Are there recognised learning pathways or recognised accreditations that people can go and do this training or that training or is there still a sort of self-taught element to it?

Micah Hoffman (07:36.756)
So I think there's both. I come from a background of cybersecurity where it was... all you had to do was stand up a system that you could attack or defend. And then you launched these attacks against it. And you learned the skills yourself or you did a capture the flag exercise online and learned a skill. OSINT is very much like that where it has these types of exercises that you can try at home. We have them on our My OSINT Training in our courses.

And there are also capture the flags where people can go and experientially learn how to do something, how to noodle through some some challenge. But then there are also more formal paths and a lot of universities, a lot of other organisations are creating those learning pathways that may lead to degrees. There are certifications out there right now, and this is my personal opinion, that right now our industry is moving

towards having better certifications, but right now a lot of the certifications are very vendor focused. So take a training class, get a cert for that company's idea of what OSINT is instead of what the industry says it is.

Matthew Stibbe (08:50.934)
It's Like doing a Cisco network engineering course rather than a network engineering course.

Micah Hoffman (08:55.474)
Yeah, exactly. So something that's very particular that says, hey, I'm a master of OSINT according to what these people say. And there's nothing wrong with that. It says I'm at least this skilled or experienced or I've gone through this training. And what I usually tell people is before entering into these programs right now, look at what you want to

where you want to go in the field of OSINT. Look at the jobs and see what they're asking for. If no one's asking for the XYZ certificate, don't invest your time or your company's time and money in getting that unless it has the knowledge you need. Right now, industry-wide certifications, they're starting to come up, but nothing is really overwhelmingly like, this is the cert that you need. or this is the...

Matthew Stibbe (09:49.15)
So There's not an equivalent of a bar exam or a CPA type of qualification. right. This is, because this as an outsider, this is, this is new to me. I don't know that. Tell me you, you, you, we've talked a little bit about My OSINT Training. Tell me about the other hat that you wear with The National Child Protection Task Force.

Micah Hoffman (09:51.764)
Thank you. Yes.

Micah Hoffman (10:07.796)
Sure. At The National Child Protection Task Force or NCPTF, that's a local,  a nonprofit over here in the United States. What I do there is I work cases. The organisation works directly with law enforcement to find missing and exploited children and to augment their law enforcement's ability to find things. Instead of law enforcement having to learn how to do things on TikTok and Snapchat and Instagram and all these other platforms,

they can focus on some of the other tasks that they're really good at. And then a team of volunteers, in our case,  the people that work the cases over at NCPTF with us. We work on these cases, create a intelligence report and then give that to law enforcement for them to action on whatever tips we were able to come up with. And that that might mean, hey, look, Johnny that's missing is posting with his friend over on TikTok in this account. You might want to look into that.

or something like that. Over there? go ahead. I was going to say, you know, over there, I Not only work cases a little bit, but I'm also a tool maker, a system maker. My, my thought is if I can help others be more successful, that's where that's where my my happy spot is. So I make systems, I work with our partners that donates lots of really great software and expertise to us as well.

Matthew Stibbe (11:10.456)
No, no please continue. I just...

Matthew Stibbe (11:35.822)
What's your experience of the law enforcement community? What's their level of understanding of OSINT and how can it help?

Micah Hoffman (11:44.756)
I think in certain areas, law enforcement has a very good grasp of open source intelligence, of social media, and we call it social media exploitation, using social media to help solve crimes or to locate people. I think law enforcement has a very good grasp in many areas. But you've experienced this too, I'm sure, in your life, that some places are more resourced than others. And for those places that...

maybe don't have the staff or don't have the monetary budget or the skills to learn those, the things that they need to get into OSINT and social media effectively, they many times may not focus on that. And that's where we fill the gap. Our nonprofit is privately funded and we, like I said, we know how to do things in social media and other places and put pieces together to help them do their jobs.

Matthew Stibbe (12:41.422)
To what extent is it capability enhancement and to what extent is it just sort of resources on the case, adding more brain power to the investigation?

Micah Hoffman (12:54.996)
It's a little bit of both. We always let law enforcement run the cases. It's their case. It's their responsibility. I think we like to think of ourselves as more lead generation. For instance, law enforcement might share a username with us or a phone number or email address or something like that, or even just the name and birthdate of the missing person.

And then we go into the software that's been donated or we'll just go right on to different social media platforms and examine what those accounts, what accounts are out there, what their family members are posting and all, and provide that. So I think for us, it really is a balance between that capability enhancement and the fact that we have tools that law enforcement can't necessarily afford or...

in some places they can't use, for like facial recognition and breach data and things like that. We can use this and they might not be able to.

Matthew Stibbe (13:59.502)
As an outsider, sometimes hear people going crowdsourcing investigations and sort going to Reddit or something, and we've got this picture of something, does anyone recognise it? And I imagine that if I was in law enforcement, that would feel a little bit sketchy, it's a one step up from a psychic phoning me up saying, but you're not, I know you're not that, but I'm just, wondering how you get past that sort of perception if it exists.

It definitely exists.

Micah Hoffman (14:28.074)
it definitely exists. No, it absolutely exists. And there are many cases or examples where well meaning people are trying to do the right thing to help do something. There was a very famous Netflix show called Don't F with Cats. I don't know if that was popular. And essentially a group of well meaning people tried to find a person that was harming animals and ultimately harming people.

But they did so in a... using open source techniques and social media techniques that are very common for other people, for us, people like us to use in OSINT. But the way they went about it, they didn't protect themselves or their systems. The person that was the, the, the person of interest in this case joined their group and was watching all of the things that they were talking about and sharing. So crowdsourcing...

We've done it for years  in the United States. We we used to put kids pictures, missing kids pictures on milk cartons. Yeah, exactly. So the crowdsourcing is out there. I think what distinguishes us from just the people on X number of social media platforms is is a reputation and the reputation for results and intelligence versus raw data. Our our director of investigations is just simply amazing at

Matthew Stibbe (15:33.428)
of milk, right?

Micah Hoffman (15:55.016)
taking all of those different data points that our volunteers and other people collect, putting it together into a meaningful, intelligent report that is solid, it is actionable, and it leads an investigator on not a wild goose chase, as a lot of tips can.

Matthew Stibbe (16:12.546)
It must be a wonderful A wonderful feeling when you've actually contributed to finding a missing or exploited child.

Micah Hoffman (16:18.31)
It is, it is, it is something that was unexpected. I come from a technical background, as I mentioned, cyber security, and I'm used to computers and, bits and bytes and things. And, when I got into the finding missing child area, it was really a shock because on the cyber security side, you know, I would hack a system legally. would hack a system for a company.

And I'd write up a report and said, here's how I got in and here's how you can fix it and go do that. And they'd be like, thanks for testing our site. And I never knew if those results were helpful. And most of them went to making somebody's bottom line a little bit more profitable. For this, we have such great rapport with a lot of the law enforcement agencies that we work with that we're able to get some feedback sometimes that little Johnny was recovered safely or

or when something else happened, we get that feedback and that is incredibly helpful for keeping us motivated and also keeping us engaged in the case.

Matthew Stibbe (17:29.634)
I want to pick up on something that you mentioned just now and when we were talking earlier. You described, let's say, amateur OSINT investigators participating in crowd investigations, not thinking about their own security and protecting themselves. Thinking more broadly about that, people in the profession, what do you think they need to be doing for their own...

operational security privacy?

Micah Hoffman (18:02.004)
We'll talk about that as kind of a sliding scale if you think about it. For instance, your house, right? Your house, you probably have one to two locks on the door. In some places, we put three or four. Maybe we put a bar across there or a chair. We wedge it inside if we're really secure. But we also need to protect our windows and we also need to protect other avenues of ingress into the house. Same thing when you're dealing with protecting yourself when you're going online.

I like to think of it as three different layers. We need to think about the network, where my traffic is coming from. If I'm using my work's computer and I'm on their network, well, some websites will say, hey, look, you're at XYZ company or XYZ government agency, and they might treat your traffic a little differently. So we need to think about your network level. Then we need to think about your system and are you protected from viruses and trojans and other things?

What information is your system giving out to those websites? And how does your profile look? And then we also have the person, you who are you when you go online? That would be the third piece. And that's a very important and kind of legally and ethically charged topic because with the network, we have virtual private networks, VPNs, which I know can be a little bit touchy nowadays, but they protect our data and they allow us to

pretend to be somewhere else. We have virtual machines at that system level or virtual appliances or virtual browsers that allow us to really mask or change or isolate what happens when we're doing our OSINT. And then for our protecting ourselves, we have virtual personas, research accounts, sock puppet accounts, which are real profiles that we maintain on social media platforms. And there's a bit of ethical

issues with that in some organisations, in some countries, creating false... creating these accounts that are not essentially Michael Hoffman's but they are you know John Doe or Jane Doe on that platform so that I can do the searches and queries and it's not attributable back to me. So it very much is like protecting your house. You have to look at all the places that data might go out or things might go out or things might come in and then take take opportunities to mitigate that.

Micah Hoffman (20:28.584)
But you also have to think about the threat against you. If you're investigating Russian oligarchs, you're gonna take one level of precaution. If you're just looking for missing children locally, you're gonna take other ones. So there's a range of things that you need to think about when you're doing this type of work.

Matthew Stibbe (20:48.662)
Yeah, we were talking about this. I was talking about this with other guests and I've had guests come on with backgrounds very carefully masked out like that. And there's a sense of which the OSINT profession can almost like, well, if I was investigating myself, what would I find and what would I want to reveal? Yeah, it's interesting. I was talking to Skip Schiphorst a week or two ago who I know you know.

Talking a little bit about how people can protect themselves emotionally from what they're investigating and is that something that... it must be an emotional burden in many ways to be looking for missing children. You you must share some anxiety about them. How do you protect yourself? How do you sort of put some professional distance on that?

Micah Hoffman (21:43.278)
It it actually affected me more than I thought it would. I've I've been able to watch things on TV and and and you know play violent video games all my life and and didn't really touch me until I remember I remember the opening scenes of the Walking Dead TV series. I don't know if you ever seen this, but there's

Matthew Stibbe (22:09.052)
hate zombie things, I  I can't watch them!

Micah Hoffman (22:11.112)
Yeah, I know exactly Right, I can't even but but you everybody talking about him like, all right, let me look at it. So in the opening scene, there's this little girl at a gas station and she gets hurt. And I was like, no, I'm not watching the rest of this series because there's something in me that that just really gets upset when kids are are are hurt or in danger. So yeah, how do I protect myself? Well,

a couple hours from now I have a meeting with my therapist. So I started seeing somebody that I can talk to about all the different things that I'm seeing. I also have a very good opportunity at work to take time off or take into access resources that can help me take a break. I've set up different environments so that when I'm doing that work on a more sensitive case,

I can isolate and segment that or I also know some positive coping strategies that I like to use when I'm in a situation where I'm stressful, where I'm stressed or where I'm impacted. You know, sometimes you think that something's going to be impactful and then it's not. And then other times you don't realise that something really hits you when you thought it was going to be fine. So...

Matthew Stibbe (23:34.062)
And it It would go off like a delayed action fuse, right? I mean, it's not necessarily in the minute. I used to be a pilot and I was very rarely in actual danger. But the few hair-raising moments I had when flying the plane, I was absolutely cool as ice when it was happening. After I landed, about an hour after my brain had landed, I was like, that happened. And it's very strange the way that the brain will... Anyway, I'm defusing myself there.

Micah Hoffman (23:56.744)
Yeah.

Micah Hoffman (24:02.514)
No, I I absolutely agree. We worked on a case and on Friday, I believe we had a debrief and the law enforcement officer told us what had happened. And I was like, well, okay. And then all weekend went by and Monday morning came by and I started working and I just couldn't stop crying. I mean, I wasn't like sobbing and stuff, but like I was, my body was just

 just crying and I'm sitting here working and I'm just crying and I was like, well, this is, I remember thinking myself, this is totally odd because I'm fine with what happened, but like something inside of me was absolutely not. And that's when I started seeing a therapist and really taking a look at my mental health as part of the taking care of Micah and doing this work long-term.

Matthew Stibbe (24:57.294)
And it sounds like you're doing all the right stuff and it sounds terribly important. I think it's a good role model and example to people out there in the profession who might be similarly affected. Let's move on from this. Let's talk about something... Well, I for one welcome my new AI overlords, but  if you're listening, there's a wonderful

Micah Hoffman (25:15.508)
So.

Matthew Stibbe (25:21.838)
poem by Richard Brautigan, I think, called All Watched Over by Machines of Loving Grace, and he wrote it in the 60s. And it is, I have this sort of idealistic and optimistic thing that we're all going to end up living in, Iain M. Banks's Culture, all watched over by machines of loving grace. If that doesn't happen, and on the way we have to use AI for open source intelligence, what are the pros and cons?

Micah Hoffman (25:28.98)
Mmm.

Micah Hoffman (25:46.58)
Well, the first thing that we always have to remember is say please and thank you. Because again, one day they're going to be able to look, they will remember and they'll be like, you're the one that always said, please, we'll give you a quick death. But I think what you have to realise about AI and again, I don't know everything about AI. I know what I use it for and what I don't use it for. I like to think of AI as like a tool.

Matthew Stibbe (25:53.868)
They will remember!

Micah Hoffman (26:13.448)
Like you'd have a hammer in a toolbox or something. Use a hammer for certain tasks is really good at those certain tasks. But when you try to use a hammer to put in a screw, well, I mean, you can hit it and it'll go in, but it's not what it was made for. When you use a hammer to pretend to try to be like a saw or something, it just doesn't work. I think right now, actually, a year ago, people were using AI for everything.

And they could, but it wasn't the right use for the tool. We are seeing AI really blossom in the capabilities from creating realistic research accounts to augmenting our abilities using tools. For me, I use AI for tools-based assistance. I know that other people are maybe using it to help collect or parse data very rapidly.

Some people are using it to analyse data or annotate data. You know, I think right now I'm less likely to use reasoning models for analysis of complicated, real world events and insights into it. I just don't have the trust there, but I also don't do that work all the time. For me, working with AI on a tool that I want to release like the bookmarklets, it's easy. I can see the results. I can...

I can challenge it when I need to challenge it and it's within my realm of understanding. I think the biggest thing is we always have to check on our AI overlords to make sure that they're doing the right thing and make sure that it's what we want them to do.

Matthew Stibbe (27:58.552)
I think one of the genius insights or accidentally clever things that OpenAI did is turn the AI into a chat because we invest quite a lot more. We fill in the gaps and we, you know, in that format, than in other formats. And it's taking a while for other formats to catch up. But I think humans are very, very good at

Micah Hoffman (28:10.206)
Mm-hmm.

Matthew Stibbe (28:26.453)
imagining intent, imagining reasoning, investing in, because we read books, we will, you know, and Bilbo Baggins doesn't exist but in our heads when we're reading Lord of the Rings there he is. And I, so I think, I think that, I think that the,

for me, the biggest danger is this sort of being suckered in by my own imagination when I'm interacting with it and thinking it knows more, thinks more. And sometimes it comes out with this enormously confident assertion of something that I actually factually know is wrong. And if I didn't know, I wouldn't know. So this sort of sanity source checking is a challenge, isn't it?

Micah Hoffman (29:05.052)
It is and there are things you can do to minimise some of that. For instance, you can ask for sources and ask for other information about where it learned type of things. The things that it's included in its reports. But again, I don't have the trust that I can have AI do a certain thing  and reason and analyse it. Although I will say that,

again, using it as a tool to do a discrete task is something that I think it really excels at. I'll give you an example. I had a missing child case that I was working on and in a photo on this person's social media account, there was a street sign clearly in view. And so I looked at the state where the person might be and I was thinking to myself, well,

I can search for all of the places where this street is or, and I went to my AI buddy and I said, hey, I want you to, in this state, find every street named this, you know, whatever, Maple Avenue, Maple Lane, Maple Street, and then I want you to make a Google Maps link in a table so that I can just click it and it'll open a new tab and then take me to that place.

and it took a second or two, but again, it's very concrete, factual information that it can grab, organise for me, and then make it easy for me to access. And I just had to...

Matthew Stibbe (30:37.334)
you've got You've got a human validation step in there because you're looking at the picture.

Micah Hoffman (30:41.33)
You got it. And you know what, about the seventh link down, the seventh Google Maps link, I was like, wait a second, this kind of looks like the area. And we were able to locate where that image was taken. But if you can think about it, is this a big task? Then I can have somebody else or AI offload and make it simpler for me to continue on my journey. I think that that's a really reasonable place for AI to help us out.

Matthew Stibbe (31:07.224)
with that this is a fascinating topic. We've got time perhaps one more question on this before we wrap up. What's the coolest thing... what's the most surprising thing that AI has done for you recently?

Micah Hoffman (31:21.672)
So it goes back to trust. When I was working with AI, initially I have a web page like Duck.ai and I'd ask it a discrete question, get a discrete response, and then I would do the work. And, you know, I started doing this more and more with a project I was working on, a project I was coding, so I would ask it, you know, code related questions, like you could do this. But it didn't see the code that I was making.

It just knew that I had that question or this question and gave me an answer. It never, it never saw the results. Well, after doing that for weeks, if not months, I decided that it would be okay to allow AI into my integrated development environment where I have actually my source code to see all of the 'Franken-code' that I had pasted together over countless conversations. And I said,

when it got in there, thinking tactically, thinking discreetly, I said analyse this code base for redundancy, for cybersecurity weakness, and optimise it for... and suggest optimisations for efficiencies. Well, it turns out that a lot of the individual questions that I had been asking and putting into my code were correct, but when you look at the entire document, when you looked at the entire code base of stuff I was working on,

it was inefficient and suboptimal. So one of the biggest 'aha' moments was when they were like, well, because of how you did all these things, Micah, here's all the problems you have. Luckily, it helped me fix it too.

Matthew Stibbe (33:03.161)
I think the big thing, coming thing, growing thing with AI right now is this, it's moving from asking a question about a thing to understanding the context and the breadth or understanding your whole code base. So when you ask it a question, it goes, yeah, you could do that, but it affects this. I upload, you know, I do these sort of...

people do these sort of psychological profile things like Myers-Briggs or whatever. Now I uploaded the last three that I'd done over the last few years into an AI thing and I said, okay, now look, this is, I recognise myself in this. Now you know this about me. And then I asked it a series of questions. And the more context it has, the better it gets. Absolutely remarkably psychologically acute recommendations came back from it.

Micah Hoffman (33:50.418)
Wow. Okay.

Matthew Stibbe (33:50.678)
So then I So I thought, I'm now just gonna say, you need to remember this about every question I ask you. So I ask a question about, I mean, I'm useless at DIY, but I ask it a question about DIY and it says, you know, given that you are an INTJ, completer, finisher, obsessive person, don't over obsess about it. Here's the simple way of, okay, so it, yeah, I think context is going to be increasingly important. And on that bombshell, I think we're pretty much out of time.

Micah Hoffman (34:06.631)
Right.

Matthew Stibbe (34:18.478)
Micah, it's been an absolute delight talking to you. Thank you so much for joining me today.

Micah Hoffman (34:22.644)
Well, thanks for having me on Matthew. I've enjoyed this time.

Matthew Stibbe (34:25.526)
And that brings this episode to a sad conclusion because I think we could keep talking for a long time. But if you'd like to learn more about OSINT, Videris, or Blackdot or you want to listen to more From the Source podcasts, please visit blackdotsolutions.com. And thank you for listening.