The developer episode or the "devisode"
Email’s Not Dead: Season 2, Episode 3
The developer episode or the devisode
Email's Not Dead
About this episode:
Hosts Jonathan Torres and Eric Trinidad dive into the emailing underworld and come back out with a distinctive look at the way developers and marketers send email. In this episode, we decided to bring in a madman. Well, we actually brought in one of our lead developers, Alan Erwin. He’s on our deliverability tools team and we wanted to ask him what made him tick. Tune into this week’s episode as Jonathan and Eric remind you, once again, that Email’s Not Dead.
Meet your presenters
Manager of the TAM team at Mailgun by Sinch
Technical Account Manager II at Mailgun by Sinch
Software Developer at Mailgun by Sinch
Email’s Not Dead – S2, E3: The developer episode or the "devisode"
Eric Trinidad: Welcome to Email’s Not Dead. My name is Eric and this is Jonathan.
Jonathan Torres: Hello.
Eric Trinidad: Hey, thanks everybody for coming back. We're continuing our series on that sweet, sweet sending funnel and going further down the pipe. But we thought we'd bring it a little bit different. Want to change it up a little bit.
Jonathan Torres: Yeah, good change of pace.
Eric Trinidad: Yeah, just something different. And we thought you'd like to see or hear well you can't really see them because you can hear his voice. With us today, we have Alan Erwin, one of our devs here at Mailgun. Welcome, sir. Thank you for being with us.
Alan Erwin: Hi, everybody. Glad to be here.
Jonathan Torres: I like that title that we just created for you, devineer.
Eric Trinidad: Devineer.
Alan Erwin: I feel a little suave right now.
Eric Trinidad: Yeah. No, I know. I saw you got a little swagger now that you did. You should have a cup with that on it for sure.
Jonathan Torres: Yeah. Devineer so you know a dev definitely not an engineering degree, but I think you're close enough.
Alan Erwin: I'm working on it.
Jonathan Torres: I think devineer at least. I like it, I like it alot.
Eric Trinidad: So we wanted to bring you in today not only to talk about some of the things you've been working on because they go kind of hand in hand. They actually are in the sending funnel that we've been talking about this past week or in the past weeks that we've had shows, but essentially wanted to get your take on kind of the backend stuff, who you are as a person. What makes you? What brought you here? All those things more.
Alan Erwin: Well, okay, so let's see. I've been doing software development for 11, almost 12 years now. I started right out of it, like in college, working at Rackspace as an intern. I got my job there, didn't know how to code and learned to make coffee. And that's what I did to keep my job.
Alan Erwin: I kept the devs that were teaching me happy by making them coffee every day and sort of just started picking up stuff and learning as fast I could to keep my job and then sort of progress from there. So I worked at Rackspace for nine years across all kinds of products there, ranging from their global authentication system to their usage system, like, just depends on their backend services, mostly. And I do the same thing here. So here I work on the validation service and inbox placement product for Mailgun.
Eric Trinidad: Yeah, right on. And the validation product was something we learned in the sending funnel that we talked about. So we'll talk a bit, a little bit about that later as well. But you know, Mailgun is a platform for developers. With that said, you know, before we get into all that, like what are some things that, like, interested you in interest... of like that you're interested in?
Alan Erwin: Oh, so just like side things that make me sort of tick?
Eric Trinidad: Yeah, yeah, yeah, yeah.
Alan Erwin: So like, I'm all over the map. Ranging from like crossfit workouts to like playing DND with folks here to like building random, like, closets that I work out of.
Eric Trinidad: Yeah? Alright.
Jonathan Torres: I think that needs a lot more explanation than just random closets to work out.
Alan Erwin: So one of the things that's really awesome about the culture here at Mailgun is that we're really incentivized to work in the conditions that make us most productive. Right. So my management has made it very clear that if we want to work remotely, we can work remotely and to make it easier for me to be remote. I have two kids, a wife and two dogs, and it's very hard to, you know, keep things quiet at my house.
Jonathan Torres: That's a house full.
Alan Erwin: Yeah, for sure. And my kids are like, like now walking and toddlers and stuff. So to make it easier for me to work, I took over my master closet and, you know, put in a desk and a little mini fridge and mounted a TV and just started working.
Jonathan Torres: That is the dream.
Eric Trinidad: Yeah I was going to say awesome and amazing but it came out "awmazing."
Jonathan Torres: If we're going to combine words today that's all we're going to do. Already gone with devineer, awmazing. So I'm sure we'll come down the line with something else. That is really, truly like a little partial dream. Like I have my own quote-unquote office in my house, but I never get to do anything fun with it. Like it's just kind of like, it's there and I'm like, I'm going to fix this, I'm going to do something more, I'm going to build this up. And never so I mean, kudos, the fact that you're taking over that and I know you're building stuff out, yet you haven't fully built it out yet.
Alan Erwin: I've got some ideas, like, my wife has approved a secret door. Yeah. I'm going to make a secret entrance into it. So that's coming probably by the end of summer... it'll be installed. So it'll be pretty awesome.
Jonathan Torres: Yeah. All right, cool. I mean, to get out of the personal stuff, some more stuff you do as a devineer. A little bit more. But I do have one question because I know you're into this stuff. Harry Potter. What is your Hogwarts house?
Alan Erwin: Okay, I'm definitely a Gryffindor person. Hands down like, the chivalry. All about it.
Jonathan Torres: All right. Awesome. Just curious. Just want to throw that out there.
Eric Trinidad: No, no. Me too.
Jonathan Torres: Coming from a Hufflepuff. You know, I can't really say too much. I know. I'm sorry. I took the test. That's what it said.
Alan Erwin: Yeah. I landed in the Gryffindor house when I took the test.
Eric Trinidad: Get out of here, Hufflepuff.
Jonathan Torres: Alright, I guess now we can move on to more stuff but it's cool, like I mean, considering all of that, like I think with each one of us, I know for me personally the stuff that I've experienced in past jobs and doing the things that I've done throughout my life also like has influenced how I work and what I do for work and things that we even do on this podcast. Right. Like life experience just kind of carries with you throughout and everything you do. So like, taking that consideration. Like, I know you've done some stuff here, like at Mailgun. What kind of stuff have you been able to work on? What kind of stuff have you done? I mean, I know... also fun fact, two year anniversary for you coming up in a couple of days.
Alan Erwin: Yeah, a couple in a couple days at the end of March 4th. Yeah it's been crazy, two years.
Jonathan Torres: Yeah. I mean, that's the fun part though, right. Like when you, when you have that experience but like. Yeah. So I mean one congrats. And two like during the time that you've been here, like what kind of stuff you've been able to do or work with or have had fun doing.
Alan Erwin: If you're talking about products, the first real project that I was sort of tasked to take over was the bulk validation feature here at Mailgun. Right. So we had a validation feature that people could use and a single API call not into their forms, things like that. And we wanted to be able to open up the idea that people can give us lists, a large list of email addresses, and we can get them validated as fast as we can. Right. So one of the things that is really cool about Mailgun is that whenever we're asked to do a task, we own it. It's ours and we can run with it however we want. And we can give designs for solutions and we get feedback. But generally, if that's the way you want to take it, you do it, right. For example, with this bulk validation feature I went with using Cassandra and using Kafka for, like, splitting up work. And it was a ton of fun, right? Just being able to really get in there and just build a solution and own it completely. A lot of big companies like Rackspace or like Microsoft, you have, like, architects that sort of give you work and help you design those kinds of solutions. Here you kind of are and all the way. And you work with the business and you work with, you know, the front end folks to build the APIs the way you need to. And so that's one of the things that's really cool. Whenever I started here I took over the bulk validation feature and it was just sort of like, run with it. And so that's one of the big starting points here at Mailgun and being able to help run the team that I'm working on and stuff like that. It's been really cool.
Jonathan Torres: That's awesome. When you start comparing it to other places like, I don't know, I've never been on that side of the house. I could never be on the side of things, like, I know what it's like to work with some support teams and things of that nature for sure. I've definitely done some tech stuff, but not really in the sense of development, more of like manual tech work, which, you know, it's cool. But I mean, all that stuff is very processed, very, you know, and to the fact that it has to be right, a lot of that stuff is very much laid out. This is a process. This is what we're going to do and how we're going to do it. And whenever there's flexibility, I love that when you can kind of be creative with some of this stuff. So, I mean, just to keep that in mind, I know you've got to talk a little bit about that. We'll surely dive into it, I'm sure, just a minute or so. But yeah, like you mentioned, the validations, like what can you tell us about the validations? Like how did that expand your journey here? What did you get to do with it? Like what was the thought process as you kind of started building some of that process out?
Alan Erwin: So when I first came on, I had very limited knowledge around email. I didn't really know what like, SMTP was and I didn't have a real idea of what our customers would need. But I understood back end services. Right. So my first foray into understanding is to get a job here, I had to basically build Mailgun, right. As a developer, say, how would I approach that? Which gave me an idea of how mail is done. So whenever I look at validations, right, we go and we do everything we can to understand risks and do the standard mail interaction just without sending data. Right. Because we're not, we don't only be a spammer, but we also want to be able to help our customers understand risk and do all that. So whenever I got to use that product… I got to build that product, I was also able to say, okay, let me talk to support. That was something really good to do at Rackspace was, hey, support. What exactly do you get back from our customers with need for validations like is there a risk. Right? Like does it, are they, are they getting the right results? And that's something that I like about Mailgun is we're encouraged to reach out to other groups, which is not necessarily the same at other companies.
Jonathan Torres: Yeah, definitely.
Eric Trinidad: Yeah, that feedback portion. But we're, you able to get a lot of feedback from the customers because I know we did beta testing for a while, you know, and I had some customers that I worked with that were involved in that beta test like was it overall well and good or even the bad feedback that you received, were you able to take that back?
Alan Erwin: Absolutely. That's one, of thing that's really great is that I had never been on a customer call before I worked here. I legitimately was always hidden back in the back end services for companies. And that was one of the things that was very cool was that as we were building out both validations, right. We had a customer that came in, who was really interested in using it. And they were a use case that we were building it for and they were willing to give us beta information and really talk and work with us on as we were honing the feature. And so the good and the bad, some of those conversations were rough, but we took some really good feedback. And the thing that's really cool about Mailgun also is that our development process is really fast. Right. So our ability to take feedback from a customer and turn around and give them what they're looking for was very short. Like the window is very tiny. So the customer is able to get a feel for that... we care about what they say. And that's something that's very different from a lot of development teams, which is really cool.
Jonathan Torres: Yeah, I think that really from my perspective. Right, I've never been a dev like I just mentioned, but I got to see in this process like, how it kind of worked and what the different steps were because we were all part of a process. Right. And yeah, to your point, whenever there's the ability to be able to communicate with everybody, ability to get feedback from a lot of different people, ability to get input even at the slightest level, like, just helps out so much. I remember when products like this started building out and teams like what you're involved with and you yourself would come around and give access to Google Docs like, you know, or things of that nature. Right. Like that. We can kind of go in there, put some information in, put some things down where we can all share this information and share the wealth of knowledge of, hey, I see that you're doing this check here for X service or, you know, maybe this should go later in the process or earlier in the process. According to what we know of when things break and when we know that there's going to be problems and being able to kind of filter that through and kind of iterate on that before a product ever went live is super amazing. One, because I've never had that experience, because we are usually exactly like you just said. Right. There's usually a product team that's filtering through to some architect that's filtering them to a development team. And you don't get to be a step of or a piece of every part of that process. And doing that and being able to see that process and be a part of that process. One is amazing and two, I think it just helps out so much being able to get a true filter of what needs to happen, how it needs to work. What are the thoughts behind every single little piece of this? Because like you said, you don't really know a whole lot about e-mail coming into this. I knew a whole lot, about too much about email coming into this. Like, I wish there were things that I didn't know about email. But, you know, like it's one of those things that, you know, with everybody working together, everybody being kind of being able to bring their piece of the puzzle into it, I think just helps out so much. So it's awesome to hear it from your side, like, how much, you know, that helped you guys are the things that you found interesting about that? And for me, like, that's definitely part of what I found interesting about it, because everybody's got input. But when everybody's doing the right input for the right thing, for the piece that they know I think is super, super helpful.
Alan Erwin: Absolutely. One of the things that's awesome is our leadership supports that, right? Yeah. Our CTO is all about like, go take the initiative and do your thing. And it's expected that we go and find the people that know the most and leverage that information because it doesn't help me at all to try and build in a silo right, because at the end of the day, it's going to have to support the product that I'm pushing out and sales is going to have to sell the product that I'm building. So not knowing what it's supposed to do and not knowing what is needed to support it doesn't help anybody. And all of our dev teams feel that way. They just, they all build things and they work closely with our support teams to fix problems as fast as possible. And we are on call for the products that we build. So it behooves us to build systems to have fast alerting in like, flash deployment and things like that, so that when support says, hey, there's this problem that needs to get fixed, we can fix it quickly or sales says, hey, we have this really cool opportunity in the market. Can we get there? And we can say, yeah, give us two weeks and we'll build something that can do that. And that's something that's very rare in companies that are fast growing like ours.
Jonathan Torres: That's awesome. Yeah, that's really cool.
Eric Trinidad: After that first call, like, like how did you feel like going into it. Like never speaking to customers before.
Alan Erwin: The first one I really didn't talk much sort of just sat there like I don't want to mess this up but I think some of our sales folks do a really good job with helping shepherding us because I'm a dev. It's very hard for me to communicate effectively sometimes with customers. And so the salesperson leads that conversation. And then whenever there's technical expertise needed, that's when I kind of step in and say, okay, this is exactly how it works. Let me give you more information. I can dig deeper. And so the next conversation, I got a feel for what the customer's depth was and what they wanted to know. And then I could have that conversation and it made it very, very easy to be able to communicate effectively with them.
Jonathan Torres: That's awesome if you ever need practice. I've actually done like the whole role playing thing. I actually let it, you know, everybody knew coming on had a role play against me. But let them know what they need to do. So if you ever want that, let me know.
Alan Erwin: Deal. Set that up for sure.
Jonathan Torres: Well kind of going along those same lines, right, once we kind of get past the development stages, once we kind of get into the actual product itself. We are very API focused, right. Mailgun itself, API centric by devs for devs is kind of been a little bit of our side motto or internal motto at the very least. Or how did that process go about? Right. Like, you've got to think of it in a sense, I'm sure, of how will this be used through an API and how is that going to come down the pipe whenever we're doing those kind of things, like anything that you find interesting or different or kind of catered toward that as you were building things out.
Alan Erwin: So whenever I go about designing an API, I kind of have to think about it in a way of if I weren't building it, what would I want? How would I want it to behave? And because I understand the fundamentals of a REST service, which is generally what we do here, to expose to our customers, it allows me to say, OK, what can I do and why can't I do? If I'm really following the idioms I will look at, I'm exposing, for example, bulk validations. We will expose the idea of a job resource and a customer would be able to say, okay, I started a job. What is it they're going to be able to do with this? And I'll talk to our front end folks and say, hey, what kind of data are you going to want to expose out to the customer through their portal? So whenever I build this API, I'll come up with a API contract and say, this is how it's going to behave. What does this feel like to you? And then that'll come back and we'll go through iterations and making sure the contract’s right for whatever feature we're building. And then we'll build a feature, test it out in staging, get some beta testers on in production and then get the feedback. And that allows us to change his contract if we need to before we go live with it. Of course, we have versioning on our API to allow for new development, for example, and validations. We have a v3 feature of a v3 release and then we have v4 one. V4 gives you more in-depth knowledge around what we're doing with validations. So, for example, you get a risk assessment in v3, you don't get that. So because we're using a rest API, we're able to add features in versions and allow customers to kind of make decisions around what behavior they want. Right. Which is really neat. Something that I really appreciate about making rest APIs is that customers are able to decide for themselves what they want, what they want to do. And for example, if we have a reseller that wants to sell validations or our inbox placement feature or something like that, integrating into that API, they just sit on top of it and say, okay, I'm going to put my own UI. I'm going to be able to, you know, have it hook into my service layer and everything, just sort of work as long as we have a really good, solid contract. So that's basically how that works. Yeah, yeah.
Jonathan Torres: But it's cool to see the thought process behind it. Right. Because not everybody, I guess, you know, everybody does everybody at some point has to think about things like, how is this going to work, how is this going to process what is everything going to be. But yeah, like it's nice that I think it's done this way, like being the whole API centric part of it and being able to do it and the ease of use of a lot of this stuff. I mean, just to get your opinion, kind of how you're processing through that, the different steps you've got to go through to build that stuff out, because we have a lot going on within Mailgun, because there is so many APIs, there's so many different things you can do with it. And, you know, one of the things that you had mentioned earlier is like, right when we were having a candid conversation about all the stuff and you had mentioned that once you put out a release for an API, you're kind of locked into that. Right. You can add certain things, but you're not able to change most of what's there for an API version unless you update and create a new version for it. So with the validation feature, I felt like there was a quick turnaround in that between the three version that does all the things that particular thing does, including being able to put it out somewhere in a public place to make the calls you need at a very quick pace. And then the v4, when you're talking about bulk, you know, validations and being able to upload and view processes in that way with the risk assessment as well. So the fact that we're or that Mailgun itself was able to iterate on that and do things at that pace and be able to do things, I don't know. For me, the way that I look at it, whenever I'm thinking like, well, I want to make sure that these addresses are correct. Both of those satisfy the needs. One for definitely for the front end side of whatever platform you're trying to create or do or or set up. And the second one is, yeah, give the tools for people to be able to do this in mass, be able to do it the proper way and be able to have something that is specifically built up for that. So rather than just trying to like, jumble something all together, the ability of being able to separate those out in the way that, that that was done, I think is pretty damn awesome, which is, I mean, one of the main reasons I think we wanted to talk to you specifically. But, you know, it's nice. It's nice to see it. Nice to get feedback from the internal side of things of how that worked.
Eric Trinidad: Agreed. He has his way of thinking and is way about and down this path, and I think from companies that we've worked for in the past, even before I met Jonathan, yes, there was a time that Jonathan and I didn't know each other. But it was almost a decade ago.
Jonathan Torres: Too long.
Eric Trinidad: Single tear. Yeah, I worked with other companies where it was just taking, like, so long between iterations and things changing and things really happen at a breakneck speed here. Like, really, things are able to get changed and updated and get the latest DLC for our validation services out there and running. So I think that's amazing that, you know, we're able to work so quickly with each other. How does your team kind of move on and think of prioritizing some of these different changes and things that are happening.
Alan Erwin: First, whenever we deliver a product? Right. There's always a support need that comes into play. Right. So the first thing is once it's delivered, we say, all right, where do we need to harden things up, make sure that they function and we've got more monitoring and disaster recovery, things like that. And then we also work with a product team that is looking at the market and saying, okay, for validations. What does it feel like? Like what does the landscape feel like? What are our competitors doing? Every company has a product team that is doing this and we work pretty closely with them. But we also come up with ideas on our own. And that's one of them is really neat. The whole validation service came out of the idea from a dev. Right. And that's also something that is really neat about Mailgun, is that I can go to my product folks and be like, hey, I have this idea and I think we should do it. This happened two days ago and Friday I went to my product folk and I went, hey, this thing could impact our customer by adding the ability to get a feel for security around validations. And they went, yes, put that into our backlog. Let's do some research around it and kind of make a decision whether or not we should spend the time investing in it. Because our team is only four people big. Our dev team is working across two products. Right. Inbox placement and validations. Right. And so that's a lot to handle. So we have to be really strict around what we are working on and what's going to make the biggest bang for our customers. And so we have with our market, with our product folks, we can do that market research and get the feel and then they bring back, hey, look, this is what we think we should do and then we communicate with them. Hey, from a technical standpoint, these are things that we have to do if we want to scale what we’re seeing our customers are really, like, loving this product. In order to scale this, we're going to have to do this specific thing. And that's we kind of have to have that tension between the two to determine, like, do we build a new feature? Do we share things up? And that's how the process kind of works, is we have meetings and conversations around what the priority is, and then from there we stack them and just get to work.
Eric Trinidad: That's awesome. And you said something there that is absolutely like I think why we enjoy working here so much is like what's the biggest bang for our customers? Like, what's going to make them, this is all for them. Like this is a tool about them, for them, you know, built by devs, for devs. You know, this is amazing. So, like, that's awesome. That's a really great way to put it.
Jonathan Torres: Yeah. No, I agree that's perfect. Like, I like being able to be part of the process and the process when you guys come up with an idea when support team member comes up with an idea, when you know customers feedback into the product team or even the product team themselves, I come up with a different idea, being able to kind of share that and spread it around and see, you know, let's evaluate, let's really think about it and then start building on it. And the fact that we have that luxury, it's pretty damn awesome. Like one for sure. And yeah, it works. I think it's worked. It's worked so far. It's brought us here and yeah I appreciate you kind of giving it a little bit of a behind the curtain little sneak peek to see what's been going on behind the scenes.
Alan Erwin: It's fun stuff. It's probably the best place I've ever worked on. I just enjoy running around and teaching people things and like I'm a white board junkie, as you guys know, I'm constantly like being able to go and just say, hey, support, we're going to do an AMA. Let's just on a Friday at two o'clock, let's just chat about validations. We do lunch and learn and things like that. And getting people to be invested in our product, in our company will help our customers. Right. So that culture I'm constantly running around trying to foster. And I see it from people all the time, which is really awesome, because at the end of the day, those will spark ideas that I'm not going to have other people see things differently. And then what ends up on our actual, like, backlog board for us to research and do work can come from anywhere in the company, outside the company. And we can then say, oh, is this going to impact our customer? This is going to make their lives easier and is it going to make our product better? Yes, let's do this. So we definitely want to foster that here at Mailgun for sure.
Eric Trinidad: I mean, two years in, like, do you see, like, what the next two years is going to bring or are you just like, do you feel like?
Alan Erwin: Is it two or ten or two, you know, like what?
Eric Trinidad: Like you're only two years in and I feel like you've done so much, you're already, like, just to, like, put your stamp on Mailgun, you know, like, do you see yourself, like, further expanding that or like is it just with this one product or do you see yourself trying to diversify more with our product?
Alan Erwin: So the way I kind of look at working for a company is I want to make the biggest impact I can. Right. So if I'm in a dev role and I'm able to make a huge impact and then it becomes a point where I'm not quite making the impact that I was, I'll shift and do what the company needs in order to make it so that I continue to make that kind of impact. Right. So in this case, if I move into, like, leading teams and doing things like that, then yeah, sure. I just want to make sure that the investment I'm making at Mailgun that I've already made, the two years I've already invested continues to be a building block to make our company better. Right. So if that means that the next step is I'm moving to being like a technical lead and moving into running teams, cool. Let's do it. As long as we're progressing as a company and we're making our customers' lives better. Yes, that's what I plan on doing right now.
Eric Trinidad: Well said.
Jonathan Torres: Anything else you want to mention? Anything you feel like you need to get off your chest right now?
Eric Trinidad: Any shout outs?
Alan Erwin: Shout out to my team. They work their ass off. And, you know, it makes it really fun to work here and, you know, even the long days and nights and stuff that we have to work, sometimes it's still a blast to work with them. And I wouldn't work anywhere else.
Eric Trinidad: Yeah. Coming to work is definitely a pleasure. And as you know, it's not all fun and games. Of course, you know, we definitely do a lot of hard work, but when you work around great people, it just makes the day go by faster and all that much more... I was going to say more better because I just hit my mic. But I mean all that much greater. Well, if you want to hear more or look at more of our validations, we do have blogs and we do have documentation on it on our website. If you have any questions, hit up your support team members, your support staff here at Mailgun. Alan, I appreciate you. You're beautiful, man. You're a deveneer and a dev engineer.
Alan Erwin: I'll take the beautiful man part. I love flattery.
Jonathan Torres: Oh, for sure. Thank you so much for, like, doing like I said, like a peek behind the curtain I think is just awesome for people to get every once in a while. Right. It's not what this podcast is meant to be. A little bit of a love fest this time a love fest for you and you guys and the work that you guys do. But, you know, really just being able to peek behind the curtain a little bit, I think it's always super beneficial for people to see. And, you know, sometimes what people want, like, I know, I like seeing that. I know, you know. Well, that's my favorite part of a documentary series. Whenever they pull the curtain back a little bit, it's always fun.
Eric Trinidad: So I appreciate you and your team. Make sure you give them a crisp high five and a just a really aggressive hug when you see him from us until next time. Thanks.