# Aaron Gotwalt
## Blog Posts
### Luxury Software
Date: 2026-02-09
I've been shipping software for 25 years. Almost all of it was terrible. That's about to change.
I've been writing software professionally since I was in high school. I've worked on almost every level that you might describe a software engineer being needed, and it's time to admit something: almost all of it was terrible. Not terrible because I'm bad at my job or worked with bad people, but terrible because good software is almost impossible to make.
I have spent months of my life just trying to get to a decent login screen. I almost lost my mind trying to make something work in Safari. One time I was trying to figure out a glaring bug and accidentally deleted the production database using other software that I wrote.
The enormous complexity of doing anything truly right means that whatever good means is almost always off the table. Instead we're stuck with continuous compromises, accumulated mistakes we don't have time to fix, and job security knowing that however bad it is, it's still better than the alternative.
There's a lot of noise about how AI's remarkable ability to write code means it's an extinction event for software engineers. I think just the opposite - this is a Cambrian moment, where everyone who used to ask for features can now participate in the awkward process of making computers do things more directly. Things that were too small or too niche are suddenly fully-featured tools built by the domain experts who need them every day. It also means that for people like me, we've got incredible tools to do the things we've always wanted to do.
**What if more software was awesome?** It'd require engineers like me to have more time to continuously rethink as needs change, rather than just compromise in order to meet deadlines. This trade is historically untenable for executives — we make money by shipping, not by rebuilding things after all. AI is the only way through this logjam.
When I can mass-produce the compromises in minutes instead of months, I get to spend my actual brain on the choices that separate software people tolerate from software people love — the animations that feel right instead of just functioning, the error message that tells you what actually happened, the thing that just works on the first try because someone cared enough to think through all the ways it wouldn't.
These choices are how you build great experiences for people. Luxury software will be made with metal, not cheap plastic parts. I can't wait.
---
### Why Doma, Why Now
Date: 2025-12-15
Smart locks and doorbells have reached their final form, which is exactly why it's time to build something new.
I was giving a tour of the Doma offices last week to a designer whose work I've admired for years. While walking him through what we've built, I found myself saying something that crystallized how I think about this company: Doma is at the intersection of three things that are extremely difficult for AI today: design, craft, and hardware.
AI is eating the world, and will soon begin to reshape our homes. Today's smart home platforms are an unimaginative starting point built on unreliable hardware for people with high frustration tolerances. Doma is building a new hardware platform to embody intelligence into the structure that you live in.
### The Readiness of the Market
iPhones look largely the same year after year because the core idea has become very refined over time. It can only be so much glass and metal, arranged as a slab. Even folding phones are essentially extensions of the same idea. I bet I can guess what the next five years of iPhone and Android devices look like.
Smart home hardware around the door — video doorbells, smart locks — were novel fifteen years ago, at the dawn of August and Ring. But where's the innovation this year? They can only evolve so far. These products have largely reached their final form factor, and yet the way we buy and install them still bears the tentative relationship we had with them at the get-go. The market wouldn't bear a Wolf or Sub-Zero equivalent doorbell back then because we didn't really know where it would go.
Now we do. Doma can exist because society has a largely shared idea of what these products are and how we use them. Everyone wants them. This consensus is the precondition for building something more ambitious.
### Making Infrastructure
What we're designing at Doma is not merely a lock or a doorbell, but a new robotic device that moves with you in and out of your house. The integration isn't incremental; it's architectural. It requires fresh ideas about how you might control and interact with an object that's trying to anticipate your next move. We're building something that understands when you're approaching, when you're leaving, when you're expecting someone.
It's one thing to sketch all this stuff out in CAD and something totally different to mass produce it in a form that can be installed in 70% of door designs sold in North America.
This is a multidisciplinary craft analogous to Swiss movement makers, who build art — but art that must function under extreme conditions and long lifespans. The Doma door needs to work reliably for fifteen or twenty years, through weather and wear and the thousand small indignities of daily use. It requires every capability of the Doma and Fuseproject teams working in concert.
### Hardware as Platform
I'm a software engineer by training, but perhaps that means less than it used to. Everything is software now — cars, light switches, faucets, and now doors. Software permeates the physical world.
Doma, as part of its broader vision, seeks to build hardware for the home with the same reliability as other things we think of as infrastructure: plumbing, appliances, walls. But this hardware is not an end to itself. It's a platform for a new kind of home intelligence — one that's able to understand the behaviors of its occupants and move from glorified remote controls into proprioceptive anticipation.
The vision is a door that doesn't wait for you to ring the doorbell. One that greets your guests, authenticates packages in transit, and does a hundred things we haven't thought of yet. That's what a platform enables.
### Why This, Why Me
I'm excited about Doma because this is what I want in my house, ultimately. Not another app-controlled deadbolt. Not another camera that sends me notifications. An integrated system that understands how my household actually moves through space, built to a higher standard than traditional consumer electronics afford.
We've gone as far as we can go with the current paradigm. Something fresh is needed. This is it.
---
### A robot house needs arms and legs
Date: 2025-05-22
doing > thinking
Smart homes have gotten good at thinking, but they're terrible at doing.
The idea of a connected house has been around for decades, first capturing two audiences: people who grew up inside a RadioShack and people with too much money. They wired up whole-house sound systems in the 90s, added motorized shades in the 2000s, security cameras a decade later, and couldn't stop evangelizing at dinner parties. Like HiFi audio and other expensive religions, the satisfaction seemed mostly in the accomplishment itself. Whether these gadgets actually improved daily life was anyone's guess.
Then something shifted. Smart locks - the August in particular - proved this technology could solve real problems by automating things you do every day: turning your deadbolt as you come and go. Nest let you adjust temperature remotely. Ring let you see who was at the door. Each solved an actual daily friction, then expanded in ways that became irreplaceable: guest access codes, energy savings when you're away, recordings of porch pirates.
Today, Google, Apple, Samsung, and every major tech company is racing to become the brain of your connected home. Want lights that turn on when you arrive and your favorite playlist to start? Easy. Want the robot vacuum to clean only the kitchen, only at night? Done. The digital orchestration is nearly flawless.
But ask your smart home to open the door and welcome guests inside? To move furniture for a party? To crack windows on a beautiful spring morning? Suddenly you're back to doing it yourself.
This is the wall every smart home hits: they can command existing devices but can't touch the physical world. They can't carry your groceries from the car, can't flip your mattress, can't water the plants you forgot about. The most useful household tasks still require hands.
Before you ask: no, I don’t think humanoid robots are the solution, at least in the near term. It’s one thing to manufacture the hardware and another thing entirely to keep them from accidentally killing your pets, shattering your dinnerware, or creeping you out at night. I think the solution is more intermediate: embedding more capabilities into the things you see and use every day.
**A robot house needs arms and legs**. That's the next frontier: smart homes that don't just think about your space, but can actually reshape it.
---
### Joining Doma
Date: 2025-04-07
I'm excited to share that I've joined Doma Home as Chief Technology Officer, working alongside founders Jason Johnson and Yves Béhar to reimagine what smart home technology can be.
I'm excited to share that I've joined Doma as Chief Technology Officer, working alongside founders Jason Johnson and Yves Béhar to reimagine what smart home technology can be.
My journey to this company started as a pandemic hobby. All of a sudden I was spending all of my time in my 1880s Victorian home. Homes like mine are accumulated more than built, and the wiring just didn't make sense. To get the switches to control the right lights in my office without ripping the wiring out of the walls, I installed a couple smart home products. Simple. Done. Better experience.
One thing led to another, and today every light, lock, switch, fan, camera, and doorbell is quietly connected without sacrificing the character of the space. No blinking LEDs or spaceship control panels - just technology that improves our living experience and gets out of the way.
This mix of an effectively timeless home and all this new technology got me thinking - what happens when we build smart home things with the same longevity as other permanent parts of our homes?
When Jason and Yves built August over a decade ago, smart home stuff was still in its "wait and see" phase. Products like the August Lock were designed to be easily removable if you changed your mind. Today, video doorbells, smart locks, and connected lights have all become permanent fixtures, but they're still built and deployed with the same tentative approach.
Doma's exploring new, more permanent ways that smart technology can fully integrate into the surfaces of the home.
Only with hindsight can I say that my career has been driven by curiosity about emerging spaces - from building CoTweet at the dawn of the social media era to cofounding Evernow as society has shifted towards improving access to quality women's healthcare during a pandemic. I get excited about places where technology meets evolving human needs.
Now with Doma, we're exploring how we interact with our homes as Apple, Samsung, Google, and others raise consumer expectations by investing billions into the space. The opportunity to help define this feels exciting and a natural extension of things I've been curious about.
I've been asking people what "luxury software" means for awhile. Nobody seems to have a good answer. I think the answer to the riddle is that the best kinds of luxuries are often quiet - the stitching in a well-made garment or the texture of its fabric. Luxury technology, if there is to be such a thing, needs to do the same - to let you feel something, even if it's hard to see on the surface.
This is why I'm particularly thrilled to get to work with Yves Béhar, a design icon whose work consistently embodies design with purpose. Every surface and angle is considered. Together with Jason, whose deep relationships in this space make all of this possible, we're building products that let the home feel, sense, and act in ways that make our lives better.
The challenges in this space are substantial. How do you build timeless products in a fast-moving space? How do you deliver hardware in an uncertain era of international trade? How do you give consumers confidence that the durable hardware they buy today will not only continue to work but actually get better over a decade of operation? There are a million other questions I have, and few easy answers, but I can't wait to tackle them with the Doma team.
I'm grateful for the journey that's brought me here - in particular, many thanks to South Park Commons for giving me space to explore over the past year - and excited to now go build something new. If you're interested in joining us, we've got open roles for firmware and software engineering, both based in San Francisco.
Onward.
---
### Sonos Might be Unfixable (But For Other Reasons)
Date: 2024-08-26
Sure, the software might be bad, but maybe the premise is ultimately flawed.
When I was twelve my parents built a new house on a hill, which they live in to this day. It was the early 90s. My dad did a lot of the work on the place, allowing him a few luxuries we might not else have been able to afford. The house was big enough that you might have guessed they’d install an intercom, but they were smart - that 1970’s technology was at this point tired and had been revealed as the gimmick it actually was. Instead, he installed speakers in four separate rooms of the house, each connected to a central distributor next to the amplifier, and put an unobtrusive volume knob somewhere on the walls. We might not have had much, but this was luxury, and music was something we cared about.
When I bought my pre-earthquake victorian in San Francisco in the early 2010s, pulling wires was out of the question, but I wanted the same experience. I went big on Sonos gear for all the obvious reasons - great audio quality, wireless connectivity, and the ability to group rooms into zones and have the same song playing in some or all of the house. This felt like magic.
Eight zones: very respectable, with a few stereo pairs and other cool accessories in the mix. I bought a network storage device and had gigabytes of my favorite music a button press away, long before Spotify and Apple Music achieved the Netflix-like domination they’ve arrived at today.
In the years since I’ve repeatedly upgraded my equipment - first to remove the outdated S1 gear that they discontinued support for a few years ago, then a new sound bar for the media room - but the way we use the system also changed. When we got a dog who was very worried about city noises, eight zones turned into eight white noise machines that ran almost 24/7. We’d watch TV in one room or another, but not bother to send the sound throughout the house.
My dog wasn’t the harbinger of doom for my Sonos usage, though: that role belongs to AirPods. My wife and I have very different media consumption habits, and part of reaching equilibrium means that she’s often listening to a podcast, an audio book or a TV show in the comfort of her ears, regardless of what room she’s in. The white noise might even keep playing in the same room. Meanwhile, I might be listening to music at my desk, but transition to my own AirPods as I head out for a run.
Playing music on Sonos in a specific room is a bit like calling a land-line telephone - its endpoint is a specific location, instead of a specific person. There are perfectly great times and places for that: on a Sunday evening while I’m cooking, I love the sound of some old piano jazz playing from across the house while I cook dinner. It’s romantic. There are also plenty of times that you want a great home theater experience. Beyond that, though, I’m not sure the promise of “seamless whole-house audio” is all that exciting outside of some narrow party situations.
So maybe it doesn’t matter whether or not Sonos ultimately fixes its software. Maybe the promise of audio in every room of your house has been solved in ways that are both better and more flexible than any number of speakers can hope to address. Maybe that’s ultimately why Sonos threw away all the brand loyalty they had in order to support some bluetooth headphones. Maybe this is why Apple ultimately didn’t buy them despite all the brand alignment in the world a few years ago: a home intercom, like multi-room audio, is a luxury from a future that’s long past.
---
### New Site
Date: 2024-07-17
A new site for a new chapter.
I looked at this site the other day and realized the last post was from 2019. This is basically typical for me. When I was a teenager I had a stack of journals that each started with a line like, "I've bought this new journal and this time I intend to write regularly". The remaining ninety nine pages were invariably empty.
The previous version of this site was built on Gatsby. The idea was just a simple static site with posts written in Markdown that could be compiled by Github Actions and hosted on Github Pages. The design was a basic template I found and further simplified. I launched it, wrote a few posts, and then disappeared on a five year voyage building Evernow.
A few weeks ago I started a project to bring this back up to date. The goals: simple, opinionated design, no ops maintenance, and no CMS. I had a few options that I explored:
- **Updating the existing Gatsby site** – I spent an afternoon trying to update all the dependencies from five years of change and never got it to clean. I gave up after trying to port the existing design into a fresh install and getting nowhere fast. I'm sure I could have gotten to a modern, typography–centric template, but nothing felt easy.
- **Hugo** – To its credit, it is _very fast_, but I'm not sure if that's a selling point that matters. I found it surprisingly difficult to perform basic tasks, such as rendering images stored in the filesystem within markdown documents. Perhaps it'd be better if I wanted to use a CMS.
- **NextJS** – It can do anything, including blogs. I've spent a lot of professional time wrestling with it, and getting started brought up memories of past frustrations. It's more than I need, and perhaps more complex than the web needs at this point.
A friend recommended I look at Astro. Mark Horn's astro-nano hit all the right design marks for me out of the gate: it's adaptive, built on top of Tailwind, has no external dependencies after compilation, and just looks great. Smart, sensible defaults everywhere.
So, with the framework and template selected, it was mostly an exercise in moving the limited content over and getting the Astro => Github Pages action up and running.
Will I write more now that I've got a new journal? Odds are low, but I'm working on it.
---
### Swimming
Date: 2019-01-28
It took me until somewhere in my Junior year at college to realize that I was just hyperventilating, that I didn't have a heart condition, and that I could run more than two miles without feeling like death. I peeled those electrocardiogram sensors off of my chest, went outside into the cold, and really haven't stopped running since. The ghosts of that time still linger -the feeling of futility during a cross country race in slushy north-eastern woodlands as a teenager is probably what prevents me from doing anything competitive to this day. I'll tell you that it's because running is for me, an act of meditation, but there's another story under the surface.
It took me until somewhere in my Junior year at college to realize that I was just hyperventilating, that I didn't have a heart condition, and that I could run more than two miles without feeling like death. I peeled those electrocardiogram sensors off of my chest, went outside into the cold, and really haven't stopped running since. The ghosts of that time still linger -
the feeling of futility during a cross country race in slushy north-eastern woodlands as a teenager is probably what prevents
me from doing anything competitive to this day. I'll tell you that it's because running is for me, an act of meditation, but
there's another story under the surface.
In second grade my school had math square dances that taught multiplication tables. The "5-10-15" song is so catchy that you'd probably fake an injury to avoid hearing it, which is exactly what I did when I was eight. "I pulled my groin," I cheerfully explained as I maneuvered my crutch to the side of the room. I don't know if Mrs. Schleppi ever caught on, but in retrospect it must have been funny, even to my weirdly conservative christian school's weirdly conservative elementary teachers. The truth: I wasn't good at dancing, and I didn't know how to fix the problem.
I didn't grow up with video games, and on the rare occasions that I end up in social situations where people are good at them, I prefer to be a spectator. I find them disorienting, and I hate the feeling of losing for lack of fundamental skills rather than game difficulty. It wasn't until buying a Nintendo that I realized that playing wasn't a sort of innate skill, but rather something that's learned and practiced over time. It seems comically obvious in hindsight.
I'd like you to think that I enjoy snowboarding, but the truth is I'm very happy to survive a few runs without any major damage and call it a day. "Text me if it's great," I'll tell you over lunch. I'm good enough to not fall over, but going fast on snow is pretty terrifying. If I think about it too hard, I'll forget to look where I'm going and wipe out instantly. Snowboarding isn't natural, at least to me, and that's how I decide if I'm good at it.
My parents sent me to swimming lessons at the local recreation center for years. On paper, it's probably the sport my body was built for - I've been long and skinny for my whole life. I'm not entirely sure of facts at this point, but I failed multiple swim tests because I couldn't do the back float, didn't like the feeling of failure, and, as far as I can tell, I quit.
If you asked me about nature vs. nurture, I could ramble on about how study after study shows that in just about every discipline talent is equally distributed, but access to opportunity is not. I could tell you about people in my life who've worked incredibly hard to overcome obstacles and become world class at something they love. I could tell you all of these things and then laugh when I consider my own path.
I ran a lot last year, and I'm feeling good about that. I'm 37 though, and starting to realize that I probably can't do this kind of high-impact exercise without real risks. There's a pool in my neighborhood full of equal parts retirees and frenetic college students. A day pass and twenty hyperventilating laps later, I realized I was still that sinking kid in my head, aware of my mediocrity and unsure of what to do next.
Left to my instincts, I'd probably just not go back. I don't like doing things I'm not naturally inclined to do. (This the same reason why I have an accountant.) This time, though, I'm trying a different approach: I got a coach for a few sessions. Though I'm probably not headed towards the Olympics any time soon, I don't feel like I'm slowly drowning anymore. I'm moving faster with less effort. I _almost_ enjoyed it earlier today.
It is the modern way to avoid discomfort, and it's not all bad. I'll probably just order delivery tonight instead of going out on a hunt for fruits and small game in the wilds of San Francisco. I can't avoid everything I'm bad at, though, and growth sometimes requires that septuagenarians cruise by you as you struggle to make it to the end of the lap. Perhaps the real exercise in swimming is in just keeping my head down and allowing my ego to let go of the goal of perfection.
---
### A Brief Memorial
Date: 2018-06-16
It’s June 16, and my grandfather’s funeral just took place on the other side of the continent. He was a kind and loving man. He outlived his siblings, his first wife (my maternal grandmother), and so many of his peers that all went to World War II together, never to be seen again. He was a child of the depression, a voice of frugality and common sense, but was the first to help me chase this weird career I’ve had.
It’s June 16, and my grandfather’s funeral just took place on the other side of the continent. He was a kind and loving man. He outlived his siblings, his first wife (my maternal grandmother), and so many of his peers that all went to World War II together, never to be seen again. He was a child of the depression, a voice of frugality and common sense, but was the first to help me chase this weird career I’ve had.
His is the last generation without color photos of their childhoods. The war is at this point abstract — some newsreels and photos, faded monuments to the fallen — dangerously forgettable, and gently drifting away as his mind did in the last few years. In his most lucid moments it was the stories of being a bomber pilot instructor that always came to the foreground. Wings dangerously close to falling off, engine failures over the water, and terrified crewmen unable to do their jobs. His brothers died and continued to die from the impact of the war’s most brutal memories. He was a lone survivor. I can’t imagine that sort of loss, or what closing his eyes must have replayed in his head.
I’ll remember him bringing me a John Deere tractor for my birthday, and the joy I felt come from him as he lifted it out of the back of his Chrystler[^1]. I’ll remember his endless fascination with clocks and archaic mechanical instruments, something that courses in my veins today. I’m not sure I ever understood him as someone who had faced great loss until he began to fade away.
I’m struck by the duality of love and loss that he endured. My family has been incredibly fortunate relative to many I know, and at his passing I think it’s important to reflect on the unseen sacrifices of those who came before us, trying to survive and make a better life for us. It challenges me to see others experiencing war and enduring deep trauma today with compassion and kindness, for we will be their children someday.
[^1]: I got into an argument with my grandomother about this. I remember it as a Cadillac, but they, and I'm quoting, "would _never_ own one of those.
---
### Shared Out
Date: 2017-06-25
I joined Twitter on July 17, 2006. I’m user #1888. I took a photo of Kevin and Mike as they were trying to keep their servers up on the first day of Instagram’s public launch (note: both have upgraded their wardrobes since). I used it when it was called Burbn. I didn’t go to an Ivy League school so I couldn’t be on the cusp of Facebook’s launch, but I knew about it at the time. I was there.
I joined Twitter on July 17, 2006. I’m [user #1888](https://twitter.com/gotwalt). I took a [photo of Kevin and Mike](https://www.instagram.com/p/CF2/?taken-by=gotwalt) as they were trying to keep their servers up on the first day of Instagram’s public launch (note: both have upgraded their wardrobes since). I used it when it was called Burbn. I didn’t go to an Ivy League school so I couldn’t be on the cusp of Facebook’s launch, but I knew about it at the time. I was there.
There was a moment (ok, fine, lots of moments, and occasionally it still happens) where I thought my tweets were clever. There was a moment where I thought I was a good Instagrammer. There was a moment where I thought that any terrible photo could get a certain number of likes just by applying the right filter.
It’s harder and harder to feel like whatever content I’m making is meaningful. Even this blog post is plagued by question as to whether it’s different than anybody else’s, and whether it really needs to be said again. Also: I have 20+ screenshots of women on Bumble with photos from Machu Picchu. I am the victim, but I’m also the plague. Maybe all of this technology, when it comes down to it, has removed any reasonable belief that we’re unique.
I find myself gravitating to Instagram stories — they’re ephemeral, and my life is pretty routinized, but even if I post similar photos of my exercise routine a couple times a week there’s little evidence to say I did.
I’m a harsh critic of my own tweets at this point. They seem to carry such weight. Who’s going to read them? Is this on brand? Will this alienate some group that I care about? Have I already posted this photo? Do I only take photos of my running route? Am I just a political ideologue on Facebook? Is there more to my digital life than this?
[Ten](https://twitter.com/gotwalt/status/236809572) [years](https://twitter.com/gotwalt/status/355536192) [ago](https://twitter.com/gotwalt/status/357881822) [I](https://twitter.com/gotwalt/status/368227912) [never](https://twitter.com/gotwalt/status/370280592) [cared](https://twitter.com/gotwalt/status/395250662) [about](https://twitter.com/gotwalt/status/409221732) [this](https://twitter.com/gotwalt/status/410760612). I did hate your Uggs though. I miss the innocence.
---
### A New Years Resolution
Date: 2017-01-03
San Francisco is one of the most amazing places in the world you can live today for tons of reasons, but after eight years here, it’s clear to me that community really isn’t one of them. This is the town where best friends try schedule dinners once a quarter, replying “yes” on an event invite just means you might think about it, and people move so frequently that you’re often better off meeting SF acquaintances in some mutually-agreed-upon destination like NYC or London.
San Francisco is one of the most amazing places in the world you can live today for tons of reasons, but after eight years here, it’s clear to me that community really isn’t one of them. This is the town where best friends try schedule dinners once a quarter, replying “yes” on an event invite just means you might think about it, and people move so frequently that you’re often better off meeting SF acquaintances in some mutually-agreed-upon destination like NYC or London.
I’ve spent a lot of time complaining about this — even as a quasi-successful tech founder with passable hygiene habits, it can just get lonely around here. I was reading [this piece](http://www.saveur.com/pableaux-johnson-red-bean-roadshow-new-orleans#page-2) about [Pableaux Johnson](https://twitter.com/pxnola)’s Monday night [red beans and rice tradition](http://cooking.nytimes.com/recipes/1016675-red-beans-and-rice) and it struck a nerve — I love to cook, I love to have random folks in my house, and nobody ever really does anything on Mondays.
Food’s a natural way to connect with people across everything that tries to divide us — politics, religion, socioeconomic status, travel schedules, mobile operating systems, whatever. I miss the after-church social fabric of my youth, and maybe this is a way to reconnect to some of the best parts of that. The food isn’t the subject as much as the conduit to connectedness.
If you’re interesting in joining me (and us), let me know. My last name is my gmail address, and I am fairly findable on the internet. This is just an experiment, but I want to figure out how to get to know my neighbors better, to listen more and talk less, and maybe share a glass of wine with people I don’t even see eye to eye with all the time. It’s the only way we’re going to ever survive.
---
### Sounds of 2016
Date: 2016-12-20
These aren’t the best songs of this year, or the only ones I’ll remember, but this collection of songs are the ones that found me at meaningful points along the way.
These aren’t the best songs of this year, or the only ones I’ll remember, but
this collection of songs are the ones that found me at meaningful points along
the way.
**Satan by D.D Dumbo**. I’ve never heard of D.D Dumbo before, and this track is
head and shoulders better than the rest of the album. There’s something
completely weird and magical about the layers of complex rhythms, rich
orchestration, and vocals you could mistake for
[Synchronicity](https://open.spotify.com/album/7yDxJXFPl88Dt9kBo0dDD6)-era
Sting. Was this the best single track of the year?
**Which Will by Amber Arcades**. Nick Drake + Smiths + Mazzy Star. I love
everything about this. [She’s also a lawyer for the international war crimes
tribunal](https://www.theguardian.com/music/2016/jun/20/amber-arcades-interview-fading-lines-annelotte-de-graaf),
which really raises the bar for artist/activists. Her album [Fading
Lines](https://open.spotify.com/album/2BmPfGhWI3bGkdWOHet6bq) is somewhere in my
top 20 for the year as well.
**The Bird by Anderson.Paak.** This year has been so turbulent for me — work,
personal life, family, everything — and this album was just a breeze of fresh
air. I love how much empty space is left in this track, and the way the bass
line sits against the beat. This is a song for the best days of summer.
**We The People… by A Tribe Called Quest.** This album and everything it
represents has to be near the top of any list for the year. The history, the
incredible influence of these people, the pointed social commentary in this
song. But that beat. Holy shit. That’s why this song is in this list. It’s just
so hard. [This may be my favorite thing that Black Sabbath had anything to do
with.](http://www.whosampled.com/sample/464188/A-Tribe-Called-Quest-We-the-People...-Black-Sabbath-Behind-the-Wall-of-Sleep/)
**Cranes in the Sky by Solange.** This song, the album it’s on, and her live
performances this year are on a totally different level than the rest of this
list. There’s the string textures, the relationship between the bass and drums,
and her voice. Then there’s the message — so intimate, so multidimensional, so
tinged with pain and injustice — [this is art that comes from the
future](http://www.thefader.com/2016/10/06/solange-a-seat-at-the-table-essay),
but it’s so much more than that.
**Drive by Lera Lynn.** Changing gears to some alt-country. This is a standout
song on a standout album full of haunting vocal performances and memorable
hooks. Whoever arranged and played the baritone guitar in this track is a
genius.
**True Love Waits by Radiohead.** It’s probably expected to put Radiohead on a
best-of list every year they release something. This year has been incredibly
painful for me as my marriage fell apart. This old song, finally recorded on an
album, has played in my head on repeat during the worst of it. [The haunting
lyrics](http://genius.com/Radiohead-true-love-waits-lyrics) and simple piano
arrangement ([reminiscent of Christoper O’Riley’s take on
things](https://open.spotify.com/track/2i9RAM4n1aAWTpWw0LRI1L)) have held me at
my lowest points this year, and I love it for that.
---
### Bay to Breakers, by the Wifi Numbers
Date: 2016-05-16
I set up a Wifi access point over Fell Street in the panhandle for Bay to Breakers, and left an open SSID for “#SfWifi”, which is used in all San Francisco parks that offer free internet access. I figured it’d be an easy one to measure traffic as people ran / walked / staggered by my house.
I set up a Wifi access point over Fell Street in the panhandle for Bay to
Breakers, and left an open SSID for “#SfWifi”, which is used in all San
Francisco parks that offer free internet access. I figured it’d be an easy one
to measure traffic as people ran / walked / staggered by my house.
Lots of iOS users. Incidentally, congrats to Matthew, Monica, Jessica, and
Jennifer for making it at least four miles into the race. If your phone is named "\'s Phone'" like mine is, you're announcing this to every network you connect to. In total, I had just
over a thousand connections from 7am until 5pm.
I wonder how much digital noise I leave running from point A to point B. For
next year, I’ll probably expose more SSIDs like “Netgear” and “Starbucks” that
lots of folks have already configured their phones to connect to. It’s also
interesting to see auto brands like Audi and Ford regularly connecting to these
access points as they drive along the street, often times exposing the names of
the driver in the process. Data is creepy.
---
## Projects
### Haptic Communicator
Date: 2016-12-20
You know no matter where we are, we're always touching by underground wires.
A long time ago an ex-girlfriend and I tried to learn morse code so we could
have side-conversations while holding hands in public. Turns out I’m not good at
that kind of signal processing task, but the idea of micro-languages stuck with
me. Maybe one hand squeeze means “let’s begin a conversation”, two means “I need
a stronger drink”, and three says “let’s bail on this party”.
Imagine a
wristband with three buttons on it — no screen or other distractions, just a
vibration motor for haptic feedback. Pair it to your phone, and have your
partner do the same with theirs. When you press button #1, their wrist vibrates.
That’s it. Give another one to your BFF and save button #3 for Mom.
With the help of my friend ([and](https://www.youtube.com/watch?v=_TAaj2Q-I-Q)
[uncommon](http://www.darwinaerospace.com/burritobomber)
[genius](https://www.youtube.com/watch?v=BwVoir5HSo4)) [John
Boiles](https://twitter.com/johnboiles), we put together a pair of prototypes.
It was fun deep dive in to the world of embedded microcontrollers, bluetooth
stacks, and physical product design. I just about destroyed my sewing machine
trying to build neoprene wristbands to house them. Ultimately, the cost of
manufacturing made it difficult to make a business case for it, but we learned a
ton in the process.
One of those things was that the gentlest, most unobtrusive notification from
the right person will get you to drop everything and pay attention. Our phones
are so intimate — going _literally_ everywhere with us—and so universally
connected that it’s easy to lose track of the signal in the noise of everyday
communications. It’s incumbent on applications to get much smarter about those
small, subtle cues and leave the sea of distractions behind.
---
### Ghosty
Date: 2014-08-01
A ghost to haunt your friends' Sonos houses.
Ghosty started as an extension of some experiments I'd been doing with the local Sonos API. Sure, you could build your own custom music playing user experience, but what else?
Ghosty was designed to make any house a haunted house. From the confines of a hidden Raspberry Pi, it picked a random time of the night, chose a random Sonos speaker from the network, played a creepy noise at low volume, and then covered it tracks. It was designed to be run for weeks without detection. Think of it as a high power, low severity prank.
Honestly it worke a bit too well, and I couldn't keep quiet about it. Good for my colleague's mental health, I suppose.
Fast Company wrote [an article about it.](https://www.fastcompany.com/3034135/hack-your-sonos-sound-system-to-terrify-your-friends)
---
## Work History
### Doma Home
CTO (2025-04-01 - Current)
Leading engineering at Doma Home, building a new hardware platform for the smart home alongside Jason Johnson and Yves Béhar. Doma sits at the intersection of three things that are extremely hard for AI today: design, craft, and hardware.
We're not building another app-controlled deadbolt. We're building integrated robotic devices that understand how a household moves through space — a door that doesn't wait for you to ring the doorbell. The engineering challenges span embedded systems, computer vision, industrial design for mass production, and building software that anticipates rather than reacts.
---
### South Park Commons
Member (2024-07-01 - 2025-01-31)
Spent six months between companies exploring ideas at SPC. Dug deep into the future of fitness, AI support for aging in place, and what a post-LLM product development workflow actually looks like in practice.
---
### Evernow
Co-Founder / CTO (2019-03-01 - 2024-04-01)
Co-founded and led engineering at Evernow for five years, building a telemedicine platform for evidence-based menopause and aging care. We grew from an idea to a team of 50+ serving thousands of patients across the US.
I built the initial product end-to-end — patient intake, provider tooling, pharmacy integrations, subscription management — and scaled the engineering team as we grew. Healthcare software is a particular kind of hard: regulatory constraints, real clinical stakes, and patients who need things to work the first time. It taught me more about building reliable systems than anything else I've done.
---
### Doctrly
Co-Founder / CTO (2017-04-01 - 2018-04-01)
Built tools for hospital security teams to automatically redact personally identifiable information from data transfers between devices and services. Healthcare data breaches are almost always preventable — Doctrly was an attempt to stop them before they happen by making redaction automatic rather than aspirational.
---
### Projector
Co-Founder / CTO (2014-11-01 - 2017-03-01)
Built a programmable, AI powered push notification platform for product and engineering teams. Projector delivered the right message to the right user at the right time using deep user modeling and personalization, built on Apache Kafka and Apache Beam, scaling to support extreme and unpredictable notification workloads from user-generated content.
---
### Seesaw
Co-Founder / CEO (2012-10-01 - 2013-12-01)
Seesaw was a consumer product designed to improve the way we make decisions with help from our friends. In addition to front-of-house responsibility, I also functioned as CTO and implemented a large portion of our back-end infrastructure.
Seesaw was an amazing learning experience — the consumer mobile app space is unbelievably fickle, and no combination of amazing investors, great App Store promotion, and top-shelf execution guarantees any amount of success. I'm proud of the work that we did, but it ultimately failed to gain traction and we sold our assets to Byliner in 2013.
---
### CoTweet
Co-Founder / Chief Architect (2008-04-01 - 2011-08-01)
Co-founded CoTweet, one of the first enterprise social media management tools. We built the platform that companies like Ford, Microsoft, and JetBlue used to adapt to the new world of public customer service — making it easy for call center staff to engage customers on Twitter at scale.
I developed the core product thesis, built the team, and led engineering, including early work on tweet attribution, delegation, and CRM integrations. Acquired by ExactTarget in 2010, which later became a foundational piece of Salesforce's Marketing Cloud.
---