Platform Engineer • Former Employee
Pros: I did a bit more than 4 years at Anaconda and saw it go through many pivots, iterations, good times and less good times.
But it has so many really bright, talented, empathetic, and kind people. It was a great place to work overall, although the experience seems to vary by team and department so I can mostly speak for engineering and the platform organization therein.
But there weren't too many meetings for engineering ICs even at a senior level, most were valuable, and we had plenty of laughs and fun while getting things done. There was truly almost no one here I didn't like on a personal level, even if a few weren't always the easiest to work with.
The unlimited PTO is not a lie, although it depends on your team, but I took off a lot of time each year and got my work done. Results are what matter on most teams.
It is fully remote. I loved that, not everyone does. Think carefully about whether this is what you want.
The on-sites aren't too common but there's at least one a year and they typically were a lot of fun, if not a bit exhausting just because of a packed agenda.
For the most part management and leadership care that it's a good place to work, give competitive and pretty fair compensation including base, options, a 10% bonus for most engineers, and lots of nice little supplementary benefits including niche things like pet insurance, fertility services, etc.
They went out of their way to keep healthcare costs down, recently, too. Employees without dependents pay very little in the US. There’s a new equity refresh program now, too, and a sabbatical for long-time employees. Flexible hours, running errands mid-day - all were possible as long as you made your calls and delivered on-time.
It's a small company so for someone early to middle in your career, it can be a source of absolutely invaluable experience. You will have to broaden your horizons enormously, just to survive. If you want a narrow lane to stick to and not go outside of, this is not a place for you. If you want clear processes and guidelines for every little thing, this is not the place. Ambiguity reigns, for better or worse. But so does autonomy, and managers mostly try to let you work on what is interesting to you, although of course that’s not always possible.
And there’s not really anyone to hold your hand, everyone has a bit of a trial by fire. Not that there is no onboarding or training - there is and a lot of effort is put into making you feel at home, comfortable, accepted- but by its nature this is a company where only really the self-motivated and proactive thrive I think, because no one is really going to tell you what to do next in clear terms typically - especially at a senior level. The fully remote makes this all the harder and punishes a lack of proactivity even more.
But for learners, doers, and the ambitious there's plenty to learn, very, very smart people to learn from, and no shortage of work to do.
It’s a place that’s diverse in people and ideas both. You will meet so many interesting people of all types, you will learn from them. In our internal surveys one thing that has been consistent is that people feel safe at Anaconda, no matter how typical or atypical, gender, race, or anything else. I think that should be a point of pride. And Anaconda has always had a “no assholes” policy that they actually enforce. I have personally seen them fire someone for just being rude/mean, and little else. Everyone cares that it’s a positive atmosphere for everyone, and does their best to embody that.
Above all despite what I say in the next section, this is a good company staffed by good people who are trying their best. It was an excellent job for me and very valuable to my career and personally. I have faith that it will come out on top, despite the obstacles and stumbles. It’s well positioned in the Python/data science/AI ecosystem, and only needs to find the focus to go further.
Cons: TDLR: The company does too much and has no easy way to do less, because everything is someone’s baby, so we can never actually kill anything off. Its open source roots and long history have left a lot of cultural, organizational, and technical debt. But, it is getting better.
Let's start at the beginning:
Anaconda has open source roots, and you have to understand what this means, because I did not when I started. It's what makes it so great and what causes so many of its chronic problems, as well. The company did not begin like a typical startup - a single idea and product that grows and branches from a single seed.
Nah, it started as a buncha kids in a trenchcoat, a series of tangentially related projects and products under one umbrella, more added with each iteration and pivot (and there were many). We were once a consulting firm, then a commercial software company, then back to our OSS roots, then a SaaS platform, then a compliance play, then a platform play.
The products/projects all somewhat related, but highly decentralized as is the nature of FOSS software, and what we made money off of was built atop it all.
So combine that with the fact that it’s a 11-12 year old “startup” of about 500 people now. For the first 7-8 years, it made basically no investment at all in infrastructure and shared services, patterns and processes. I’d find a machine in AWS that hadn’t been touched since 2017 but was burning money for the last 8 years and no one even noticed. Great for Bezos, awful for us.
So for much of my time there, it was more cleaning what I’ll call the sins of the past - which I’m proud to say in many ways we did. We made huge progress here and I’m proud of that and of my team and others who did that work.
Anaconda is now at the cutting edge tech-wise, with a modern Kubernetes stack heavily integrated into the CNCF ecosystem with observability and solid CI/CD, modern frontend frameworks and a modern Python backend. The skill of the engineers was never in question, just their ability to swim the same direction.
But despite our wins, right up until the day I left, it always felt like we were reactive and not proactive, never able to level up our platform/developer experience capabilities like we wanted because of said chronic underinvestment/drowning with firefighting work. Never able to address ancient tech debt in our software, because it was always a new thing to do instead.
So why? Why could we never get ahead of it? Because at the end of the day, Anaconda does too much, has always done too much, and continues to do too much. What it lacked above all was focus. Dozens of teams with just a few people each (sometimes just 1-2 people), toiling away in their little silo. My own team was told “this is the top priority” over and over, only to drown under a deluge of random asks and tasks unrelated to our “top priority”, most of which were never mentioned in planning once even when teams were explicitly asked to call out shared service dependencies. This was as sure as the sun rises in the morning in my 4 years there - how to deploy crap was always, always an afterthought.
There was a lack of accountability in cutting products and honestly people who were not working out. An often overly-conflict-adverse management in general who didn’t wanna make tough and painful but necessary decisions. Some in management/leadership who made bad strategic decisions but were never held accountable for those, either.
We bought a couple of other smaller companies without a clear plan for how to monetize them or integrate them, meaning we basically bought tech debt and less focus for no visible gain. Again, just a general tradition of poor foresight and chasing shiny things with no proof that there’s a market fit for us there.
I think a lot of this again goes back to our OSS roots, and the engineering culture that came out of it. A wild-west engineering culture of “every engineers solves their own problems in whatever way they can make work”. If you’ve worked in this industry long, you know that may work for 20, 50, even 100 employee…but not many more. And having to slap restrictions, revoke permissions, and clean up all this crap was very painful and still ongoing, and often pissed off engineers who were used to having the freedom to spread tech debt far and wide, in the name of innovation or expediency.
Leadership has tried many kinds of top-down planning, even more reorgs and other attempts to focus us. But I maintain that the key issue is the inertia of the engineering culture which still does not prioritize simplicity, predictability, standards and patterns. Projects only move forward by way of individual relationships and the effort of a few, while others sometimes kinda coast under the radar. Yet high performers are also historically not really rewarded in any meaningful way, either.
Again, there have been improvements. Things feel more focused than they did before. The relationship between engineering and product used to barely exist at all - engineers just kinda built what they thought customers wanted (we never knew, product never told us because they didn’t know either). What was top priority was whatever some manager yelled the loudest about, a battle royale. This is still a big problem, but investment in the product org has yielded some improvement, although not fast enough IMO. To this day, a lot of senior engineers kinda have to wear a product manager hat, too.
So I’ll wrap up with this: Anaconda is a company old enough to be entering high school soon, but is still trying to figure out what it wants to be. I do think it will get there, and the current CTO (CPTO) and new CEO give me some faith that it can, but it’s going to be hard and I know that they are very aware of that fact, as is anyone who’s been there long enough.