Transcript
Announcer: Hello, and welcome to Screaming in the Cloud with your host, Cloud Economist Corey Quinn. This weekly show features conversations with people doing interesting work in the world of Cloud, thoughtful commentary on the state of the technical world, and ridiculous titles for which Corey refuses to apologize. This is Screaming in the Cloud.

Corey: Welcome to Screaming in the Cloud. I'm Corey Quinn. I'm joined this week by longtime, I guess, friend/nemesis/I-don't-even-know-anymore, Ana Visneski, who at least in title these days, if we go by business cards, is the senior director of communications and community at H2O.ai. Ana, welcome to the show.

Ana: Thank you, Corey, it's nice to finally get to join you.

Corey: So, at the beginning, let's start with where you are now: H2O.ai. What is that? By just reading off of the tin, the .ai domain tells me that it's machine-learning powered, and the H2O portion tells me that it's machine learning only watered down. Close to accurate? Not so much?

Ana: Yes, the AI part is accurate, but the reason for the H2O part is actually more of the idea that our founder and CEO, Sri, really wants AI to be accessible to everyone, and as transparent as possible. So, a lot of times when you're working with AI, a lot of what goes into an AI application is in this big black box that no one knows what's inside of it. And the whole idea behind H2O.ai is democratize that, to make it transparent, to make it easy to access, and to make it available to anyone and everyone who needs it or wants to use it.

Corey: People who need to use it or want to use it, but do you take a position on whether people should use it? Because lord knows I do, usually cynical.

Ana: I know. We've had many conversations on Twitter about that. I think that it has applications across the board, but it needs to be used carefully. It needs to be used with the mindset of looking for bias, looking for problematic use. One of the things we do is we are very careful to try and hunt down bias within AI and make sure that if it's being used, it's being used the right way. I think AI could genuinely help the world. As you know, my background is a lot of disaster response. I think AI could genuinely help the world. But as with any good tool, it's double-edged, and it needs to be used carefully.

Corey: Sidestepping the issue of bias in AI entirely, let's instead talk about your background because I've been trying to get you on this show for ages. Your background is fascinating. You spent an inordinate amount of time in the United States Coast Guard. Then you went to AWS where first you ran launch operations for a while and then transitioned to a different team running disaster response. The fact that those two are not the same department gives me some pause and I have a lot of crappy comments around that, but tell me a little bit about, first, why leave the Coast Guard after you were in as long as you were? And then, why AWS?

Ana: So, the Coast Guard, it was just time to go. I love the Coast Guard. I'm third generation. I specialized in crisis communications, disaster response, and search and rescue. A little known fact is I was actually the founder of the Coast Guard social media program, the first official blogger for any armed service, and generally a digital-native pain in their ass.

So, there came a point with the Coast Guard where, as the chief of digital media at headquarters, I’d kind of run out of things to do besides going back to search and rescue, and it was time to go. And then why AWS? Well, I went to grad school with Jeff Barr, actually. And so when Jeff posted about needing someone to help him with the blog, I reached out to him and applied for the job, and went through Amazon's very intense interview process, and ended up coming aboard to help Jeff run the blog. That was my first job was the senior director of digital media for the blog.

And why AWS? It looked super exciting. Obviously, getting the chance to work with Jeff was just amazing, and it kind of tied together a lot of the things I had loved building for the Coast Guard into something I could do as a civilian.

Corey: You were instrumental in my sarcastic birthday video for Jeff, where we redid the tune, “Piano Man” to, “Blogging Man.” You were credited at the end of that as Barr Raiser—with two Rs—because of course you were. And thank you for that.

Ana: You're very welcome. That was fun.

Corey: Everyone has asked, what did Ana do with that? The honest answer is that you knew Jeff super well, and I wanted to surprise him with the video, but I didn't want to go too far in a direction that he would inadvertently find insulting or highlighted something he didn't like. It's all fun and games until you realize that you've basically just inadvertently insulted one of the nicest people in tech. So, that was really the core of what I needed your help with—

Ana: Yep.

Corey: —and you delivered admirably.

Ana: Yeah. Jeff, hands down, is—I would go beyond one of the nicest people in tech. He's one of the most genuinely kind, and thoughtful human beings on the planet. I love that guy.

Corey: I also want to point something out now that it is years later, and neither of us has anything to fear or anything to gain. I got to know you when you were hoping to do a lot of the blog stuff and then transitioned into launch operations, helping handle the orchestration of all the various AWS releases. And although we always maintained a good relationship, never once did you tell me anything that was not public knowledge, that was relevant or germane to a release. You never even pointed at—“Go check this thing over there and tell me what you see.”

Some folks have. You never have. You always were the consummate professional on this, which is fine. I mean, my goal was not and has never been, to beat AWS to the punch at announcing or releasing something. I’d like to make sure it is very clearly in public before I talk about it, otherwise, no one's going to believe that I didn't break NDAs. Every once in a while, I wind up periodically getting angry emails, “You just tweeted about this thing that's under NDA. How could you?” It was, “Well, I didn't know about it until I just read it on the official blog.”

Ana: [laugh].

Corey: “Oh, I guess we released it.” Yeah. No kidding. I'm not here telling other people's secrets. Why would I? There's no benefit to that.

Ana: Yeah. I used to joke that Ariel Kelman, who was the VP at the time, knew my background in disaster response and in the Coast Guard, that I had a top-secret clearance. And so he's like, “She'd be great to run launch.” So, I actually went from just running the one blog to being the head of launch blog and podcast, which is, again, kind of similar to my current title: taking a whole bunch of things that have nothing to do with each other and cramming them into one role.

But yeah, I did. I ended up running launch operations, and that's when you—I believe you tweeted something about waterboarding a VP, and one of my jobs was literally to make sure that things didn't leak, and to make sure that the launch process went well. And for anyone who's seen an Andy Jassy keynote, everything he's talking about on stage, there's a second room where we are making things go live as he speaks, which is intense, to say the least.

Corey: That's right. Yeah, I did a tweet thread on you one day, and my exact tweet quote was—and, like, I used to see early leaks, like documentation push too soon, pricing API updates, new endpoints coming online before they were announced. And one day that all stopped. And I first started to see why while waterboarding an AWS SVP trying to get roadmap information out of them. “I can’t.” He sobbed. “Ana will kill me.”

That was mostly sarcastic, but there was an element of truth to that. It was very clear that what you ended up doing was fixing a number of internal processes. You brought discipline to a previously organic process that, honestly, the organization had very clearly from the outside, outgrown. That was my perception of it, and I'm sure curious as to how accurate that assessment really is.

Ana: So, what's interesting is how fast a company like AWS grows. And like the company I'm at now, H2O, they're growing. And at some point, you have to put in a process to scale. Because you can't keep doing everything manually; you can't keep doing everything seat of your pants. But you also can't put a process in place that is so rigid that when things are naturally chaotic, you can't flex.

So, what I mean by that is that, yeah, I put a process in place around launch, and that's kind of my thing, is looking at chaos, untangling it, and figuring out how to build a process around it. But you can't build something so rigid, that whatever was working before goes away. Does that make sense?

Corey: It absolutely does. And it's easy to forget from the outside because we see our AWS bills getting bigger, we see them releasing a bunch of nonsense that isn't ever going to help us directly or personally, but what we don't see is the fact that every person we used to know at AWS is now at least 10 people as the organization has grown. And the processes that make sense, fail to scale. I mean, enterprise is itself a skill. When you're interviewing someone, “So, what is your nature of your relationship, historically, with the sales division?” And if the response is, “Oh, you mean, Steve?” That's not really the right answer. It's understanding the interplay, organizationally, is super challenging. At AWS more so than most.

Ana: Yes.

Corey: I feel like half my job is still introducing Amazonians to one another.

Ana: [laugh]. That's true. I actually had someone figure out who I was because you had tweeted about me, leaving launch. It was kind of funny— and someone at AWS. So, without talking about AWS too much because it kind of feels like talking about an ex-boyfriend.

You know, I don't want to sound obsessive. But the thing is, is that I think there's an aspect to it, to what I do, that every company can benefit from, and it actually comes from my background as a veteran. And I think this is something a lot of companies miss out on or don't think about when they're looking at veterans as potential employees. A lot of times they look at us for what did we specifically do. And I have friends who are aviators and friends who are boat drivers.

Well, who's going to need a helicopter pilot in tech, right? At the same time, that helicopter pilot understands how to triage operations quickly, understands how to make processional decisions. Like there's all of these things we do that are built into how we're trained and how we learn to process information that is fairly different from the civilian sector. And so honestly, a lot of what I did for launch operations, a lot of what I'm doing now, and I'll tell you, a lot of what I did during 2020 and with disaster response for AWS, was literally taking the same processes I used in a command center to make sure that the operations of that command center were flowing smoothly, and I translated them into corporate and used the same methodology. And it worked really well.

Corey: It's curious to me that skills that can be sharpened in one discipline apply so well to other problem domains while there's still so much of an initial resistance to bringing people in who do not have deep expertise in the problem domain they're focusing on now. It continually baffles me. It's, “Oh, we're looking for someone who's been experienced to”—I don't know, in your case—“watering down AI”—which I know is not what you do. Don't @ me—and has been doing it for at least five years.

It's well, great. You haven't been but the thing that you have been doing very clearly lends itself directly to your current role. And it requires a little bit of vision to find people who have skills that may directly translate, and I guess on some of the willing to, quote-unquote, “Take a chance,” not that it's a big gamble. And oh, wow, it turns out that you don't need to know everything about our business when we hire you.

Ana: It was interesting. When I first got to AWS, there was someone where when I was originally putting in a process around Jeff Barr, I actually won the Marketing Newcomer of the Year award in my first year at AWS, and the big joke was because I played guard dog and protected Jeff Barr. And one of the things I did was I built a process around getting in touch with him and an actual ticketing process for blog posts. And someone hadn't put in any tickets and I went into their office and I said, “Hey, we're not going to be able to do these. They need tickets.”

And I don't remember how the whole conversation went, but the thing that stands out to me is I made a joke going, “Well, maybe it's my military background that I like things to be organized.” And this person said back to me, “Well, then you're not going to fit in very well around here.” Now, fast forward to re:Invent, and said person needed me to figure out how to create two blog posts—This was my very first re:Invent, so what 2016— they needed to figure out how to create two blog posts by the end of a keynote.

And I did it. And you know what I used? That military organization skill. And if you think I was too above it to look at the person and say, “So, how do you like those military organization skills now?” You think I'm a better person than I am because I, of course, did the I-told-you-so dance. But it is interesting that, as a veteran, you run into some really interesting, preconceived notions about what your skills are and aren’t, and what you can and can't do. And that has been an eye-opener for me, joining the civilian sector.

Corey: Now, never having served myself, for a variety of excellent reasons, including they didn't want me when I attempted to enlist when I was 18 years old. There's a lot of, I think, preconceptions people have about those who have served and what it means to be in the military. My dad was a Naval Academy grad, so it was pretty clear from the age of five onward that I was never going to live up to the lofty expectations he held for me. So, when I look at it, it's through the—I have a bunch of nostalgic stories that he would tell, and a whole bunch of myths and falsehoods from those who didn't serve, or the stories of those who did, filtered through same. It's surprising to me that, first, there's this expectation that, “Oh, if you were in the military, clearly, you're only ever good at following orders and not solving anything independently.” It's anyone who's ever worked with a large organization knows that that is provably untrue. It's an easy, funny trope to make, but let's not kid ourselves. If that's all it was, then we would have a very different society than we do now.

Ana: Well, yeah. And so it's always going to depend also— there is a bit of the personality and what did you do when you were in. I can speak for the Coast Guard in that the Coast Guard as a service is smaller than the New York Police Department. There's 39,000—give or take—total in the whole country—or in the whole service, at any given point in time on active duty. That's not a lot of people.

And so that means that we were all expected to kind of be a jack of all trades in some way. You know, as an officer, yeah, my operational specialty was search and rescue but my staff specialty was crisis comms and, basically, public relations. And then because we didn't have anyone to do it and I was the one who had been doing it the longest in my own life, I also ended up as the social media guru who launched all of that. So, there's a lot of stuff that people don't necessarily understand. And you're right.

We do get kind of lumped in is we're not going to be good decision-makers, or we're going to be rigid decision-makers, that we're all very rigid, that we're all very hard, that we all are gun-toting. That was one of my favorites is, I must be into guns if I was in the military. Yes, I know how to shoot them. No, I don't own any of them. I was in the Coast Guard. I liked saving people.

Corey: Oh, the one that always resonated with me, it was, “Oh, your ex-military. You must manage by yelling at people.” It's a common misconception that that's not how military people manage. That's how Israelis manage.

Ana: Well, you did get your yellers. But you also have to figure what most people see of quote-unquote, “Military” is what they see on TV or in the movies. And they're not going to make it look as civilized as it is because it's not as entertaining. When you get a chance, watch the movie The Guardian, have a few beers, call me, and I will explain that entire movie to you.

But we don't yell. And if we yell, it's because someone is literally dying. And that was one of the things that I know, Ariel and my managers at AWS appreciated, and Read, my manager now, appreciates is, I don't flip out. Like… I mean, honestly, once you've been in charge of trying to find someone before they drown, or there's literal fires around you, or I was a responder in Katrina, once you've been a responder to a lot of that stuff, it's pretty hard for you to get spun up about anything else.

Corey: Yeah, it's hard to look at this through a lens of—how do I put this politely—giving a crap when historically, the risk was, oh, people will very possibly die. Whereas then you take it to civilian life, and it's, “Oh, heavens. If this doesn't go correctly, then fewer people might be able to view ads for the next 15 minutes.” And I'm not trying to crap on ad tech’s business models—

Ana: [laugh].

Corey: —but it's also not exactly life or death. One thing I've always appreciated about ex-military is they've never seemed to get too worked up about various outages. I understand I'm stereotyping wildly here, but it's, “Oh, the site is down. Let's go through the runbook. Let's be calm and collected.” And that's really what you need when everything's on fire.

Ana: Well. I could tell you, anyone who's worked with me knows I get fired up.

Corey: Anyone who’s ever worked with me knows that I get fired.

Ana: [laugh]. That's true. I've heard the stories. I will say one interesting thing on the benefit side of being a veteran, is that the fact that veterans quote-unquote, “Yell” or are intense, has helped me a lot in some ways because then when I'm being intense, or I'm being really driven, or I'm fired up about something, it's because I'm a veteran, not because I'm a moody woman.

I've basically been able to use one stereotype to fend off another. Which is a horrible thing to think about when you think about the way it is to be a woman in tech. But I know for a fact, I've probably been able to accomplish things because people assume my personality is based on my training, not just because I've been this way since I was a kid.

Corey: Yeah, it's fun. I find that for better or worse, an awful lot of the preconceptions that people have are, “Oh, it must be this experience you went through that makes you this way.” And not the inverse of, “Oh, being this way, naturally caused you to gravitate towards that experience.” It's a chicken versus egg question.

Ana: It’s true. It is a chicken versus egg question. And for me, I had only been in the Coast Guard, gosh, not even a full year when Hurricane Katrina happened. And I was deployed to help there, and I ended up as the federal on-scene coordinator’s public information officer. And that right there changed the path of my life.

But I wouldn't have pushed to be deployed if I didn't have the personality I have. But honestly, Katrina changed the path, in the sense that in my first year in the Coast Guard, I got to see what a difference we could make if we could communicate to the public where they could find water, where safe haven was. If we could communicate to everyone the best way to help. And I realized that I had a unique skill set because I was digital. I had been blogging for years, I was one of the LiveJournal early adopters.

Boy, that ages me. You know, all of this. And it changed how I viewed what I wanted to do within the Coast Guard. But that actually ends up impacting, too, why I left AWS. This need to be a part of something that helps and a part of something bigger than me.

And I used to joke all the time that leaving the Coast Guard was kind of leaving the Justice League. I didn't have my cape anymore. Who was I? So, that part was pretty interesting. But yeah, it does. It's a chicken or the egg. It's Katrina changed my life, but I wouldn't have been there if it wasn't for my personality and my drive.

Corey: One of the saddest things on some level is when we have something that is transformative like that, in the context of a disaster or other form of crisis, how infrequently it seems that that changes anything. But there are so many lessons there if you care to go after them, and it bugs me every time I see an organization, at least visibly, failing to learn from those things.

Ana: You might have seen on Twitter, I just talked about this, that there's a list of things I really hope people learn from 2020. And when I say people, I specifically mean managers and companies. I'm not even going to get into the government side of it; I could rant about that for three years. But at the end of the day, there's a thing called a black swan event, and you have to understand what a black swan event is. It's basically this idea that no matter how much you plan, you can't actually plan for the exact next event.

Yeah, we can look at what happened in COVID-19, we can look at what happened in 2020, we can try to plan to deal with 2020. But in 2021, something else is going to happen, and it's going to come from a different direction. There will never be another Hurricane Katrina, a Deepwater Horizon, a Fukushima event. Every event is at its core different. So, you have to understand that.

But you should also take a look at what we did that was successful in 2020. What did we do that helped? What did we do that made our businesses successful, that helped our people not burn out? Find those things because those, you can still apply. Even though the disaster might be different, the methodologies can be the same.

Corey: And I think that's sort of the overarching question that I have for you, which is, you've gone from handling blog posts to handling all of launch operations to handling disaster response and multiple companies. What's the common thread?

Ana: The common thread is that at the end of the day, they're all common, if that makes sense. So, what I mean by that is, when I'm running launch operations, each launch is an event. Each launch has a beginning, a middle, and an end. There are things that go wrong. There are things that don't work. Well, when you have a disaster, there's a beginning, a middle, and an end.

Corey: One of the hard parts, of course, is here recording this mid-pandemic. I wish I had your optimism that this disaster will end. And I'm sure that's a common thing for folks in the middle of… trauma, for lack of a better term. And it always does end, but it's hard to wrap my head around that right now and feel that emotionally, which is part of the reason we have experts in these things to guide us through it. Ideally.

Ana: At the end of the day, COVID has been a global disaster. If you look at the number of people who have passed away, who didn't need to if we had locked down the right way, and all of that, I could go on for hours, about the way this crisis has been handled or mishandled. But at the end of the day, I look for the hope in it. There are some great things that have happened, too. And on the days when it just seems like there is no light at the end of the tunnel, and when I'm looking at, I have friends who have gotten it, and I've lost friends, and it's just—it feels to impossible to carry for another day, I look and think, “Okay, well, we were able to find this vaccine faster than ever.” And that to me is amazing. Last year was literally the hardest year of my life. That's including running a re:Invent where I launched 106 net new services and/or features. Okay.

Corey: We've changed yours, incidentally. So, it was two years ago, if we're not—

Ana: [laugh].

Corey: —if we’re being very hon—I assume. You're talking about re:Invent 2019

Ana: ’18 was my last one.

Corey: Oh, okay.

Ana: ’19, remember, I actually got to see you for coffee because I was there and not—

Corey: I just figured you were playing hooky.

Ana: [laugh]. Are you kidding? I think it was. I should have worn an ankle bracelet; they would have electrocuted me. But at the end of the day, now what we need to look at is, okay, what worked last year? We can do a lot more online digitally than we thought.

One of the last things I did at AWS was helped our technical teams and helped AWS build out a good disaster response plan moving forward, including command center style watch rotations for teams so that no one burned out, workloads were more evenly spread, so that if someone does get sick, they don't feel guilty—because, of course, in our culture, you never let anyone else take your work because then you're redundant, right? Well, redundancy is important to survive in a pandemic or in a crisis. So, we need to start looking at those things. And I think one of the things that can give us hope is, look how far we've already survived this? I'm not going to say we've succeeded, or we're survivors yet, but look how far we've already come. And—

Corey: Well, those of us who have survived.

Ana: That is true. Those of us who have survived. And, again, don't get me started on the numbers, Corey because I will go off. But we do have a vaccine, and totally barring the lack of response here in the US by the administration, let's just talk about the vaccine from a technical perspective. That vaccine as you know, AWS, one of the things my team did last year was we had established the Diagnostics Development Initiative, with $20 million going to universities, and hospitals, and scientists to try and help speed up finding a vaccine.

Well, if you look, the vaccine is here, largely because of AI. And largely because of the cloud computing power that was given out and that everyone came together to build. It's not dissimilar to efforts in previous pandemics or previous diseases where everyone put differences aside and came together to try and find the best possible vaccine. And that was a really cool thing to see. So, I think as a responder, the one thing I've learned throughout my entire career is, look for the helpers.

Look for the people who want to get into it with you and just dive in, and dig in, and find ways to help. And so for you, Corey, maybe one way to help with seeing the light at the end of the tunnel is find somewhere to volunteer your expertise. I know you do a lot of volunteering with nonprofits to help them with their AWS bills. Well, maybe volunteer and target nonprofits that deal with suicide prevention and domestic violence response because right now we're seeing spikes in both of those due to the lockdown.

Corey: For better or worse. Most of the nonprofits I speak to in those spaces don't have appreciable cloud bills on some level, which makes sense. I would argue that in many cases, a company does not need a massive cloud bill. And if they have one that's indicative that something's gone wrong somewhere. But your point is well taken.

Ana: Well, and even if they don't need the billing help, as some of them are starting to get into working with AWS, they don't even know how to figure out what their bill says. So, there's ways. And to anyone listening, there's ways for them to help too. There's great hackathons going on, there's ways you can help them home, there are—in some countries they need help—there's mapathons because as they're trying to get the vaccine disseminated, there are certain areas of the world where maps haven't been updated enough to figure out what's a house versus a warehouse versus a school. So, help with the mapathons. That can help people in those areas get the vaccines to the homes they need to be at. So, there's a plethora of ways to help. You just have to look for him.

Corey: So, last question, before I wind up calling this an episode because I figure after this one, you're going to want to get the head start, running. Why did you leave AWS?

Ana: A number of reasons. I kind of alluded to a little bit of it earlier in that 2020 was a very hard year for me. And I don't think enough of us talk about burnout. And I have been through 12 years of the Coast Guard, including night watches, and basically every major disaster response from two thousand—what—four to sixteen. And 2020 was a brutal year.

And I was on 18-hour days. It was starting to physically take a toll on me. So, I decided to take a step back and I took a leave of absence. And during that leave of absence, I really thought about it. I had done all the big stuff I wanted to do with AWS: I had fixed the blog platform, I had built a disaster response program. What did I want to do next?

And I was tired, and I was not giving AWS or my manager—who I adore, I still love her— or Andy, the best of me because I was so tired. And then at the end of the day, the other part of it for me was, remember how I said being in the Coast Guard was kind of like being in the Justice League. At my core, I love helping people I love building, and building the new thing, and making technology accessible and usable to those who might not have it already or not know how to use it. That’s what I did the whole time I was in the Coast Guard was taking these new—at the time, which again makes me sound old—technologies like Twitter, and getting them into the hands where we could actually use it appropriately. And H2O offered me that opportunity.

It's a chance to help a company that genuinely wants to use AI to better and to use AI as a tool in a way that will genuinely help people. Our founder actually started it largely because his mother had cancer and he was having trouble getting data to work with to help with her situation. She is in remission and is doing well, but the whole reason the company was founded was about helping people. And so for me, leaving AWS and coming to H2O, was a little bit more of feeling like I was back to where I was happiest: using a new technology to help in new and in engaging, and amazing ways. Plus getting to build things because it is a younger company; it is a startup.

Being able to dive in and build things from the ground up because you asked earlier what the commonality was between everything I've done. Every time I've taken a new role and every job I've done, there has been a problem that I needed to fix and a system I needed to build. Every time. And it's exciting, so the reason I left was I needed that excitement and I needed that mission. And I was just damn tired. [laugh].

Corey: I can definitely understand that. Ana, thank you for taking the time to tolerate my slings, arrows, and difficult questions. If people want to learn more about what you're up to and follow along on your grand adventure, where can they find you?

Ana: The easiest place is probably on Twitter at @acvisneski. Hard to spell but, you know, I'm sure they'll find me. But that's probably the best place to do it. My Instagram is just my dogs.

Corey: Excellent. I should follow you on Instagram is what I'm taking away from this.

Ana: [laugh]. Yeah, I have very cute dogs. [laugh].

Corey: Ana Visneski, senior director of communications and community at H2O.ai. I'm Cloud Economist Corey Quinn and this is Screaming in the Cloud. If you've enjoyed this podcast, please leave a five-star review on your podcast platform of choice, whereas if you’ve despised this podcast, please leave a five-star review on your podcast platform of choice and an insulting comment about the military.

Announcer: This has been this week’s episode of Screaming in the Cloud. You can also find more Corey at screaminginthecloud.com, or wherever fine snark is sold.

This has been a HumblePod production. Stay humble.