Brian: PodRocket is sponsored by LogRocket, a front-end-monitoring and product analytics solution. Don't know what that is? Go to logrocket.com. Thanks. Ben: Hey, everyone. Welcome to the LogRocket Podcast. Today we're doing a Founder Friday episode and I'm super excited to have Madeleine Reese, the founder and CEO of Allma on the show. How are you doing, Madeleine? Madeleine Reese: I'm doing well. How about yourself? Ben: I am doing well. The weather is nice outside, so nothing to complain about. I know you and I, we've spoken a bunch over the years as fellow founders building dev tools companies in Boston, but would love for you to introduce yourself and talk a bit about Allma for our audience. Madeleine Reese: Yeah, absolutely. I'm excited to be here. As you noted, I am the founder and CEO of Allma. Allma is at a high-level UI-less incident collaboration. We are building next-generation alerting incident response and service cataloging natively in Slack. Our focus is very much on helping the entire organization to route and resolve incidents faster and be able to communicate more effectively during an incident. The way that we do this is by bringing everything related to an incident into one place, that place being Slack today, making it clear who's doing what and keeping folks up to date with what is happening throughout an incident, that way teams can focus on mitigation and resolution. Ben: I'm curious to hear a bit about your starting story. What inspired you to start at Allma and go down this journey? Madeleine Reese: For me, it was a long non-linear journey, all centered on a lifetime fascination of understanding complex systems and how to solve problems within them. My fascination probably originated with the fact that I have something called "synesthesia," which is where your senses are fused, so essentially, you'll have one sense activated and then another unrelated sense will be activated at the same time. It can manifest in different ways. Some people taste sound. For me, it manifests particularly in something called "time-spatial synesthesia" and "grapheme-color synesthesia." Madeleine Reese: Time-spatial is a form of synesthesia where I experience units of time like hours or. days or months as an abstract concept and as something that occupies very specific locations in space relative to me. For instance, I perceive a year as shaped like a elliptical hopscotch and I start in the September block and then I can visualize different months across the year. Then color-grapheme synesthesia essentially means that I see letters and numbers in color, so elementary school was a particularly weird time for me because I'd be answering questions like, "Five plus five equals yellow 10." Madeleine Reese: But because of my synesthesia, I've always been fascinated by understanding complex systems with all of their interdependencies and specifically connecting seemingly unrelated, disparate components of the system and pretty unique ways to solve problems and to understand what's happening in the system. That very much powered my education, my career, everything that I've done. Madeleine Reese: Then back in the day, a few years out of undergrad, I was early at a startup in New York where we ended up growing quite quickly, tripling in size in a little over a year-and-a-half. While it was exciting for me, it was also pretty scary. First and foremost, just the sheer number of incidents multiplied in frequency and complexity, and then broadly speaking, the system complexity of figuring out how to route and resolve all different types of organizational incidents to the right person, to the right team became so much more difficult when the system compounds in complexity. Suddenly, engineering team is now a department with a capital E and what used to be the simple question of, "Who do I pass along this customer bug report to?" becomes a pretty complex workflow. Madeleine Reese: Just generally, working through problems together as a company becomes much harder, everything from a marketing language snafu to even communicating if the light bulbs are out in a certain meeting room, a product trade-off discussion, people issues. For me, that was a huge moment of learning, just the sheer difficulty of working through incidents at scale as the complexity of your system and your software is massively compounding. It very much ignited for me a quest that I've been on ever since. Madeleine Reese: After going to grad school, I spent some time at a hedge fund called Bridgewater where I had an amazing opportunity to work for the founder, Ray Dalio, and the CEOs, and worked for them to route and resolve and root cause all different types of organizational-wide people and process incidents. I did so using their internal software that they had built for this purpose and it was really there where I was exposed to this data-driven way of using software to understand your organization as a complex system and understand your software as a complex system and apply systems thinking as a framework to route and resolve all different types of incidents and root cause them and compound those learnings so that you're consistently evolving the design of your system, the people within it, and the supporting workflows and tooling. Madeleine Reese: My time at Bridgewater really helped me understand that incidents draw from the complexity of your software, your many different systems of microservices, as well as the complexity of humans as complex, irrational, emotional entities with brains, and that ultimately, incidents are breaks across these systems that cause sub-optimal outcomes, so mitigation and resolution comes down to examining your systems and the components within them, identifying what you don't know, making connections among seemingly disparate components to understand what's happening and to figure out what to do about it to fix it, and then afterwards, sitting down to figure out why that thing occurred in the first place so that you can learn from it and make design changes. Madeleine Reese: It was really at Bridgewater that I became inspired and thought back to myself at my early startup experience and said, "Hey, I wonder, knowing what I know now, can I help fast-growing teams be able to route and resolve incidents as they scale and as their software systems and organizational systems become increasingly complex?" That's very much what we're helping teams do at Allma. It also comes out of our team's collective experience, shared and differing perspectives, navigating complex systems and working through incidents across a variety of different roles and companies. Ben: That's awesome. Madeleine Reese: It's probably worth defining what I mean by "incidents" and "incident collaboration." When I talk about incidents, I'm speaking about any event that disrupts your operations and is causing sub-optimal outcomes. It can be a break, a bug, a defect, something taking place in your systems, your software, your services, or applications. Incidents are events that frontend developers are actually increasingly finding themselves involved in. I think there are a few forces at play to explain this change, but I'm happy to walk through them, and in doing so, define what I mean by "incident collaboration." Ben: Yeah, sure. That'd be great. Madeleine Reese: Cool. I think it's worth going back before we go forward to explain this change and how software development has been changing to trace this evolution. Back in the day of incident management, incidents were by and large managed by a very focused ops team who were exclusively the ones dedicated to mitigating and resolving incidents and also worth noting you're by and large building the most monolithic software, so you have a dedicated ops team responsible for navigating these incidents. Cue war room with ops folks hard at work. Madeleine Reese: As we've been changing the way we develop software, incidents have actually evolved alongside this, notably changes such as the expansion of microservices, the accelerated adoption of containers and Kubernetes, the emergence of strong ownership team models where engineering teams are more and more owning both the development and the maintenance of what they're building, remote teams, the explosion of dev tools. All of these changes have given us way more autonomy and flexibility, but they've also increased the complexity of managing our systems, as there's so many different layers, the different levels of abstraction. Then you layer on incidents, which are particularly chaotic and stressful and by and large manual and mission-critical. Essentially, incidents are at their core a massive human collaboration event. The reality is there's something that the entire organization is increasingly becoming involved in across engineering, customer-facing teams, exact product marketing. Madeleine Reese: If we break down these various teams and the roles that they play during an incident, you have frontend engineers taking on more full-stack roles and helping problem-solve and mitigate and resolve incidents, your customer success teams who are serving as the first line of defense with customers, communicating out updates, routing ad-hoc bug reports, your security teams, which are dealing with the data security of privacy incidents, product teams doing incident triaging, tracking, AI's coming out of the incident as it relates to the products they're managing, finance teams navigating SLEs related to customers impacted by incidents alongside legal teams. Sales teams is wondering, "Can I pull up this demo right now, or should I hold off because our website's down?" We have marketing teams which are debating what language to push or put on hold during an incident. Madeleine Reese: Because of all of these interdependencies across the organization, we need to shift the way that we actually define and conduct our team's incident process to take this into account from Old World incident management exclusive ops team to today, incident collaboration where really the entire organization is working together, collaborating on routing and resolving these incidents. We made that a primary goal of ours at Allma to ensure that from the beginning, we're building a product that really helps everyone in the company work through incidents. Ben: Cool. Take me through the process with Allma. Let's pretend we're LogRocket, so we have engineering, we have customer success, sales, all of these stakeholders that you mentioned. Take me through the process if we were using Allma when an incident does happen. How does Allma work? How does it help my team collaborate around the incident and get through all the steps of triaging to eventually solving the incident? Madeleine Reese: I will say at the highest level, our goal is to be able to help teams route and resolve their incidents faster and communicate more easily across the organization during an incident. There are essentially three product principles that have guided how we approach building the early product. I'll start there and then I'll walk through how they actually map to the flow today. First and foremost, we try as much as possible to work with teams' existing tooling and environment. We ourselves had a bit of tool fatigue, as I alluded to earlier, and so "Not another tool" is a little bit of a mantra of ours. That's where the notion of UI-less comes in, so as much as possible, we plug in and integrate with tools and systems that teams already use, Slack-native integrations with tools like PagerDuty and soon to be Jira, Opsgenie, Zoom, and more. Madeleine Reese: The second piece that we think a lot about as it relates to incidents is putting everything related to an incident in one place to mitigate context-switching as much as possible, so anything we can do to aggregate and synthesize components, events, data across a team's systems and serve that up to a team during an incident, we want to do because it helps folks problem-solve more effectively. That also includes taking incident best-practice guidance and putting it straight into the tool rather than forcing someone to go into an external Wiki. Madeleine Reese: Then the final thing we think a lot about is emphasizing the human collaboration piece of incidents that I was talking about earlier, so really facilitating communication across the organization and taking that burden off engineering teams as much as we can, making it clear who's doing what, what's happening, keeping folks up to date. Very explicitly, the way that Allma operates today is we, by and large, work in a Slack app. Everything that's related to in-the-moment mitigation and resolution happens in a Slack app. Madeleine Reese: That's because, by and large, many, many teams are already there mitigating and resolving incidents, so an engineer would use Allma to declare an incident, either manually, through a slash command, or off of a PagerDuty alert that we've brought into Slack. Then we would kick off a series of incident workflows complete with best-practice guidance along the way that guides the team through the incident process and helps them do things like calibrate severity levels, select roles, bring the right people into the incident channel, spin up that incident channel, maintain the data trail, everything we can to bring incidents in one place and streamline those workflows so that the team can purely focus on mitigation and resolution throughout an incident. Ben: Got it. Then does it also notify the other stakeholders? I think you mentioned customer success or sales. Do those notifications happen in Slack as well, in other Slack channels that those stakeholders are spending their time in? Madeleine Reese: Yeah, that's exactly right. We've also built a communication system recognizing that communications are an essential part of working through an incident. You need to be able to talk to people about what is going on, but it can be really hard to do that effectively, especially when incidents are happening in off-hours and folks are scattered everywhere. Our communications system is built within our Slack app and essentially enables teams to define the pieces of data and events related to the incident that they want to push out across their workspace to various channels, so you can say, "Hey, every time I declare an incident, I want customer team channel, engineering channel, an exact channel to receive those notifications." We have conditions as well, so you can say, "Hey, I actually just want the exact channel to receive our highest-priority incidents and that is it." Then Allma automatically generates those status updates throughout an incident, so an engineer doesn't need to remember to manually text the CTO, or a customer success person doesn't need to spend time dropping into a channel. We'll bring that information to that relevant person. Ben: I'm curious, if the tool is truly UI-less, how do you build out the workflows and actually define the workflows that people go through when declaring an incident and who is notified? Do you build those out in Slack as well? Or is there an admin UI for configuring the workflows and then the day-to-day process of triaging incidents all happens in Slack? Madeleine Reese: Yeah, so by and large, that's already built and preloaded directly into Slack. We're at this cool place with Allma where I would say, by and large, we're a little bit at the nascent frontier of true Slack-native applications and it's been pretty cool for us because it's offered us the opportunity to work closely with the Slack team in building Allma and then also helping them evolve their platform as well. But essentially, as much as possible, we have preloaded those incident workflows, embedded best-practice guidance straight into the tool, integrated directly with a team's systems and tools so that we plug into a team's existing environment and incident process pretty fluidly. Madeleine Reese: I'll note that we, of course, enable configurability along the way. To your point about settings configuration, that is something that a team today can configure their own settings, their own logic and definitions for things like severity levels, roles, et cetera, directly in our Allma app in Slack. We also do have a web app, it's worth noting, that takes into account things that are not in-the-moment incident mitigation and resolution related, so think data storage, timeline generation, analytics, and will also enable settings configuration there as well, mostly in service of giving teams the flexibility as much as possible. Ben: You mentioned integrations with PagerDuty and Opsgenie, those tools have some workflows built around incident management, but I'm curious, how do teams leverage those tools today, and then specifically, where's Allma better and why you should choose to layer Allma on top of your existing PagerDuty or Opsgenie? Madeleine Reese: I would say the philosophy we've taken is to partner with those tools and enhance what they are wonderful at, and also what teams by and large turn to them for. For tools like PagerDuty and Opsgenie, they're incredible at sending an alert reliably to the right person at the right team, which is a feat. Madeleine Reese: What we do is we take that power of PagerDuty, for instance, and we take that reliance that many, many engineering teams have on PagerDuty to do that, and then we enhance that functionality by doing things like bringing paging directly into Slack so that a team can actually declare an incident in an Allma Slack modal and view all of their active PagerDuty alerts and easily bind them to the incident. We enhance the functionality that PagerDuty offers to teams and then we really see Allma as coming into this white space of serving the purpose of facilitating incident collaboration and as much as possible streamlining incident workflows, enabling folks to communicate out during an incident, serve up relevant data events and information across a team's many different systems, PagerDuty, Opsgenie, down the line, many other tools and systems that teams use today to understand how their systems and software and services are performing and meaningfully expose that to enable in-the-moment mitigation and resolution of an incident. That's really where our focus is. Ben: It certainly sounds like you add not only better collaboration tools on top of PagerDuty and Opsgenie, but that piece about looping in the other stakeholders who are not necessarily in engineering, customer success, and sales executives. I think that's super important because I've seen this at LogRocket when we've had, we don't have a lot of incidents, but occasionally, like any engineering org, we have incidents and making sure that everyone is in the loop understands the issue, the scope of the issue, who it's affecting, and the process to resolve is so important. Madeleine Reese: Yeah, that's exactly right. Essentially, two things worth noting to synthesize what you were alluding to, first off, we really take into account from the very beginning this notion that incidents involves the entire organization and from the beginning, we built a product that, while our primary user is engineers, really serves every team across the organization. Madeleine Reese: Then the second piece of it as well is that we connect through our notion of embedding across multiple systems. We connect PagerDuty, which is a standalone tool, to many other tools and systems that are across the company's workspace so that they can get a more full picture of what is happening across their incidents and across the people working together to solve them. Ben: I want to talk a bit about the business side. I'm curious, with a product like this, what's your approach to acquiring users? Madeleine Reese: Yeah, I'll give you the lay of the land of what we do and where we are today and then I'll walk through how we're thinking about it going forward. I'll say that we're, by and large, still pretty early in our journey and we've been running a early program alongside design partners who we have been lucky enough to build Allma alongside and learn from in our earliest days. We have worked with them to ensure that we were empathetically understanding and listening to their pain and needs versus just building whatever we want to be building. That's been incredibly helpful in ensuring that our product actually meets teams' needs. Madeleine Reese: I will say going forward, the way that we think about customer acquisition broadly is essentially through the lens of how we can best serve and support our customers and community is a very big piece of that. One thing that we focus and spend a lot of time on is community. We do a lot today, even given how early we are, we've invested a good amount already. Right now, we bring our community together in many different ways to foster shared learning and collaboration to acquire customers and to support these customers. Madeleine Reese: First and foremost, as I noted, we worked very closely with our early design partners and we do whatever we can to support them, so things like we have Slack channels with all of our customers so that we can support easily and directly and gather their feedback as well. We also hold office hours with them to ensure that we're consistently learning and evolving the product alongside them and providing them support as they embark and continue to grow their own incident practices. Madeleine Reese: We support our broader community via Discord, via special edition game sessions that bring all different types of folks together. We host monthly events like our Battle Scars and Beers event, where we swap memorable incident stories over whatever beverage of choice you have. We actually just released an interview series on our digital publication called Incidentally where we share the stories of engineering and reliability leaders, founders, and around the secrets and tools that they use to scale their system and teams, so everything that we can do to foster sharing of knowledge and learning and collaboration is very aligned with our DNA, the DNA of our product, and how we look to acquire and support customers. Ben: Those design partners, those are early users who are giving you a ton of feedback and involved in the product development process. Is that right? Madeleine Reese: Yeah, that's exactly right. Ben: What was the process like to raise your first round of funding? Madeleine Reese: I would say at a high level, this is my first time fundraising for a company as a first-time founder. I really didn't know what to expect going into it. It's always that you don't know what you don't know. When I started, transparently, I was intimidated by this process that I had heard so many people talk about with fear and intimidation and pain and then I started getting into it and I realized a few things. First is that it's really an opportunity to share and triangulate your unique vision for the world that you have and your story and your passion with other people and ideally learn for them along the process, which is actually something that I really enjoy and love, so I ended up loving fundraising. Madeleine Reese: The other thing I realized that is it's quite similar to building your team and hiring and that it comes down to mutually learning about the other person and finding that overlap around shared vision, shared core operating principles, and capabilities. Allma was lucky enough to be able to be selective on who we brought on as our investor partners, and so I chose and evaluated very much, as I would with a team member around shared vision, shared core operating principles, capabilities that I was looking for in a partner and the same qualities that I look for in a team, i.e. folks who lift me up, who challenged my thinking, who offer it with transparency, who hold a high-quality bar, who hold me accountable, who make me better, I looked for and found in our investor partners. Ben: I'm curious, when you think about competitive risk when building a product like this, obviously PagerDuty and I guess Opsgenie is now, they're owned by Atlassian, I think. They're elephants in the room and I'm curious how you think. You're clearly building functionality that they don't have, but kind of in their wheelhouse, so how do you think long-term about maintaining competitive advantage when there are big companies also in your space? Madeleine Reese: Yeah, of course. I think it rests on a few different things. I think data is a very big piece of what we're interested in doing and helping teams with and because we've taken the approach of being as UI-less as possible and connecting and bringing together so many different tools and systems, essentially Allma is creating a system of action where we're taking all of these different systems of records, bringing all of this information, these tools that don't normally speak to each other and being able to synthesize that information and serve it up to a team during an incident to help guide them through that process faster and more effectively to help them make better data-driven decisions about what is happening with their software and their systems. I think that's a pretty unique approach and it's one that, in service of your question, certainly creates mode and differentiation from our perspective. Madeleine Reese: I also take the perspective of creating a strong brand and community beyond your product, being just critical to building any company today, but particularly a technology company today, and so not only do we invest a ton in our community because we learn so much for them and we want to do everything we can to help them learn, but people are ultimately at the core of any company in any product, in any business, so therefore, by putting humans first and by doing everything we can to serve our community and help them, it very much furthers our product and Allma as a whole. Madeleine Reese: The TLDR is we're doing four things differently right off the bat. The first is that we bring a unique Slack-native UI-less approach of plugging right into teams' existing toolings and environment versus Old World is very much standalone software tools. New World, in my opinion, is being able to bring information and work to where teams are already working à la consumer apps like DoorDash, Uber, Instagram, where they take a product or take a stream of information or a stream of visuals and serve it up to the user directly to where they are. That's the approach that we've taken for Allma. Madeleine Reese: The second thing is opinionated and configurable workflows that codify best-practice guidance. That's a difference as well as we're moving those from Wiki manuals and just putting them straight into the product like I've discussed. Madeleine Reese: Then third is really creating and owning the category of incident collaboration as I defined earlier and recognizing how the world of software has changed and how incidents have changed and evolved with it and ensuring that we're building a product that recognizes and pioneers those changes. Madeleine Reese: Then finally, this vision of a very much being a cross-collaborative tool for the entire company to be able to route and resolve their incidents together is quite a different standpoint and perspective that we bring. Ben: Tell me a bit about what you're excited about on your product roadmap. Madeleine Reese: Yeah, absolutely. I would say at a high level, we're at an exciting inflection point at Allma where we've confirmed our early hypotheses, and as I said, we've been working alongside wonderful design partners who've meaningfully influenced our products. We wouldn't be where we are today without these design partners and we've gotten Allma to a place where customers actively adopt Allma as their default incident collaboration tool, and so I'm very much excited at a high level to be able to share more broadly Allma with the world and just get Allma in the hands of many more teams to be able to help them work through incidents, and of course, ourselves as a team continuing to learn and evolve alongside our customers, because of course we don't know way more than we know today. Madeleine Reese: Underneath that, we have pretty exciting plans for 2021 across product community and the team. As it relates to product, we're working right now on an upcoming product functionality that I cannot wait to be releasing over the coming months. We've gotten a ton of great feedback from our design partners and we're incorporating those changes into our Slack and web apps. I would say our explicit focus as it relates to product remains on executing our North Star and helping teams collaborate on incidents faster and more effectively. The work that we're doing today by and large maps to those original three product principles that I described, so things like continuing to add more integrations to further plug right into team systems and work directly with tools that they're already using, things like continuing to mitigate contact switching by serving up more and different helpful information during an incident, and to continue to guide teams along the way to aid with problem-solving, and finally, more that we can do around facilitating collaboration by continuing to enhance our communication system for companies. Ben: More broadly outside of Allma's roadmap, what are you most excited about in web development? Madeleine Reese: Yeah, so one thing I am quite excited about is a tool that I've been using called GraphCMS. That's enabled me as a non-front-end developer to be able to build out our marketing website. I've been using GraphCMS, which is a graph QL-native API CMS to be able to build our website. I've even been built an interactive Slack emulator within the website, which folks should check out because it's pretty cool and I'm pretty proud of it, but I've done it all using GraphCMS. Essentially, the way that it's worked is our engineering team created the initial individual component like LEGO pieces and then I was able to go in and build the logic and the content and combine them and be able to take a ton of work and time off of our engineering team and also not have to bother them every time I go into build a new page or update content or add more logic to our emulator, so I highly recommend that. It's just been awesome at Allma. Ben: Cool. Yeah, no, we'll definitely put a link to that in the episode description for anyone that wants to check it out. Madeleine, thank you so much for joining us on the episode. It's been great to learn about Allma. If anyone out there wants to check it out, it is A-L-L-M-A dot io, allma.io, to learn more about the product, and we'll put up link to it and the description as well. You guys are also hiring, right? Madeleine Reese: That's correct. Yes, we are hiring and growing our engineering team right now, looking for full-stack, backend, and for engineers. Ben: Awesome. Well, thanks so much. Madeleine Reese: Thanks so much, Ben. Brian: Hi. Thanks for listening. Please remember to like, subscribe, email me if you want, even though none of you do. Go to logrocket.com and try it out. It's free to try. Then it costs money, but yeah, we'll see you next time. Thanks.