Why I chose *not* to do a startup again
As many will tell you, building and running a company is hard. You really have to be mentally prepared to do it. The first time around, I was young, not married, didn’t have a lot of responsibilities in my personal life and it was still extremely difficult and stressful.
COVID hit in early 2020
I was laid off around the end of March.
Before that, I was stressed, not sleeping well, and not happy with where I was in my personal or professional life.
As a software engineer, I have spent the bulk of my career in product management. There are some companies that I had a lot of fun working at. The people, especially engineers, that I have met at each company, have all been great to work with.
I have been working since I was 15. Three jobs in the summer before starting college. When I left India, to attend college in the US at 16, I didn’t know a soul in this country (search YouTube for that story). I had three jobs while going to college full time. Since I didn’t grow up here, I had no clue about scholarships or grants until I got to college. I was able to get grants to pay for some of it and took out loans which left me with a hefty debt. Since I graduated, I have worked pretty much full time until around the end of March 2020.
For the first time in my life, I decided not to look for a job and use that time to think. I started questioning everything about my life. What was my purpose here? What do I care about? Even working in tech, it is extremely difficult to have a middle class life here. I am very lucky though that I have great friends and live in this beautiful city, despite all it’s problems.
I have volunteered at non-profits to make this city a better place to live and work. I started a non-profit community to get residents engaged and hosted my first @joinClubhouse event on the future of San Francisco’s Housing.. more to come. This city is my home.
Over the last few years, the process of discovery of who I am and what I want out of life, which had been slow-going until 2020. Maybe because I finally took time to think. It forced me to examine my life. After my Mom died (not due to COVID) in late June, not long after her 90th birthday, things began to crystallize in my mind. She was not happy for many years with her life, and since retirement, said that she is just “passing her time.” It was very sad to see her like that. She was explicit that she didn’t want any of us to do anything to prolong her suffering and we should let her go. I’m glad I got to spend the final days with her. She was coherent and responsive for about a day and half and then she was gone.
While I grieved, late October is when things started to click. A bunch of us went to Tahoe. We rented the Scandinavian house (my wife is part Swedish), chilled out for a few days. Other than my Mom, I wasn’t close to my siblings or my Dad, who had also passed a few years earlier.
Should I do a startup again?
I realized after that trip that the most important thing wasn’t a job or a startup or money or status or power. What was important were the people in my life, the relationships, particularly, with my wife and close friends.
This is what I doubled down on. I also joined a lot of community forums to meet new people. Participated in lunch clubs online, got interested in what was happening to my city and what I can do about it. I read a lot, kept up my fitness routine, picked up activities like learning to play guitar and went back to re-kindling previous activities that I had stopped — because I had the time.
While I was figuring out what I wanted to do next, I had already started tinkering to solve optimization problems in the cloud compute space.
I had been germinating an idea. When I was working at Mesosphere (now D2iQ), in a couple of hallway conversations with @TobiKnaup and Edward Hsu I think; I remember just shooting the breeze about what it would be like if we could tap into all the idle compute that’s distributed around the world that probably isn’t fully utilized. I was thinking of all the little devices that have spare compute from Nest thermostats on up. At the time, I didn’t formally propose it or worked on it, it was just an idea. What I do remember was that the technology to do this back in 2016–2017 wasn’t baked enough.
When I was at Weaveworks (one of the best places to work, Alexis is a great person), I started formulating it into a hypothesis about how we optimize the compute capacity that the businesses were spending money on, in the public cloud. I was having a lot more hallway and lunch conversations about this. It just sat in my head and I started writing everything down. I remember having conversations about this during Kubeconwith a bunch of Weaveworks engineers, grabbing lunch or dinner. People like @Ilya, Sam, @Miklos et al. By the time I left Weaveworks, I had the idea fully formed along with the business model. It was October 2018 when I synthesized it into a one pager and sent it to a person I respected at @a16z, Ben Horowitz. I narrowed the focus to the Kubernetes ecosystem even though it was still early but maturing fast. It was *good enough*. Kubernetes workloads were growing. Kubecon attendance was increasing. CNCF was the great shepherd of Kubenetes and all its ecosystem components.
There’s at least $100B of wealth transfer that occurs annually, from businesses and individuals to the 3 largest cloud providers in a $370+ billion cloud computing marking (IDG 2019). While this is not cheap, for most companies, it’s still cheap(er) and less of a hassle than building and managing one’s own infrastructure.
How do I enable a shift in the balance of power away from these oligopolies to the consumers of cloud compute resources? How do I democratize cloud compute?
Ben’s response at the time:
“I think the problem is a real problem, but the solution has been attempted and hasn’t met great success so far. I know our team looked at some of the companies in the space. I forgot the names, but can chase them down.”
I knew that the problem was not only real but solvable where before it wasn’t. The reason: Kubernetes, which is a common abstraction layer across cloud providers and now has become the de-facto operating system of the cloud, across providers. Even if the amount of Kubernetes spend is only 30% of the $100+ billion, that was a big enough market to focus on.
In July of 2020, about a month after Mom had died, I sent Ben a progress update.
By now, I had created the website, built a mockup, talked to some prospects and experts in the Kubernetes community. I had recruited some engineers to help me part time. They were between jobs. I am so grateful to all of them who spent time with me to help me build something. I want to particularly thank Max Kaplan, Brad Beam, Alberto Llamas, Tiffany Jaya (UX/UI designer), for the early prototypes and internal MVP along with Mark Tse and Lucio Campanelli who also helped with the design.
Crypto and Blockchain weren’t in the picture at this time even though a16z is a big proponent of it. I was listening to their podcasts on crypto as I was curious about the space.
Ben introduced me to Satish at his firm. We exchanged emails, I sent a pitch deck. Nothing came of it. I wasn’t expecting anything at that time. My main issue was validating if this will be a viable business. I was convinced that the problems were solvable. I had recruited two Data Scientists to be advisors. My friend Anoop (Director of Engineering, Kubernetes) who worked with me at Sysdig was also super helpful by asking me the difficult questions and making me think about the viability. I know we could solve the problems and make money, if we could recruit the right team. I had a bunch of engineers lined up. But I couldn’t afford to pay them. Erik (@eirkhollensbe) was introduced to me by my friend @Ilya and he was super helpful. He even built the authentication MVP in a weekend out of the goodness of his heart. I’m so grateful to all the talented people that gave their time.
I reached out to my network to see if I could raise seed money. After lots of conversations with angels and VC’s, I had narrowed down to a few investors who were willing to put in money but none wanted to be the lead, so I started to look for who would be a lead investor.
Note: While I knew a lot of VC’s, I had never raised any money from them. This whole process was completely new to me. In my first company, I didn’t raise any money. I was able to grow organically. However, this one would take some serious engineering talent — experienced in distributed systems and with golang, Kubernetes, and/or container experience, and Prometheus operations.
Some of the VCs, Angels, I pitched to — all were gracious with their time and feedback
a16z (Satish Talluri), Lightspeed (Guru Chahal), RTP.vc (Gyan Kapur), EssenceVC (Tim Chen), Florian Leibert (ex-CEO/Founder of Mesosphere), Intel Capital (Andrew Fligel), Costanoavc (Greg Sands), Firstround (Todd Jackson), Mangocap (Robin Vasan)
Other experts that provided an ear and advice:
Max and I met and we hit it off so we formed a partnership and became co-founders. Unfortunately, he couldn’t afford to not be making money because of the lack of healthcare, which I completely understand. He was great and did a lot of the operational work and also built out one of the core MVP features, an automated ASG using virtual node pools. We now had two of the core features that we could go out into the market with. We were still missing a couple more. By this time, we had lined up 13 prospects and one was willing to pay if we could deliver something to them to try out in Q4 2020. I started to look for people to work as contractors to do this but this would cost money and by now I have to think about how I am going to make a living and not go into depleting my savings.
In late summer, Brad (one of the early people on the team) had been working for a startup and he told the founder and CEO about what I was working on and thought they would be synergies since that company was also a ML company trying to solve optimization problems. There were a bunch of startups already in this space starting to pop up. This company had recently raised about $63M ($70+m total) in a series B round. We kicked off the dialog between us. By the fall, I had made progress and I reached out to Matt (Co-founder and CEO) who was in the middle of an acquisition at the time. They emerged from that as a re-branded company. Since I didn’t have any money to pay people and I didn’t have a job, what should I do?
I knew I had something and I knew what my personal motivation for doing it is. By now, I had also figured out my philosophy of how I would build and run a Product org and the company. I was clear in what mattered to me and what didn’t. This was a freeing experience.
I started interviewing in the late fall.
All the interviews were enjoyable since I wasn’t trying to prove anything. I am comfortable with who I am and what I am capable of. I was completely transparent with both the recruiters and the companies about my thinking. I had seen many examples of what worked and didn’t work in all the companies I worked at. I reached out to @SachinRekhi who was gracious in his feedback on my resume. He asked me a question that I had to think about — thanks Sachin! Do I want to be an entrepreneur? Or, do I want to work for someone else and will I be happy there? I had to think about this.
2020 gave me time to reflect. I realized that building a startup is a means to an end but what was the end? I needed to answer *why* I want to do what I am doing?. What do I want to work on if I am to spend a ⅓ of my daily life on it at a minimum. I felt that sacrificing my relationships and friendships by not having time for them is not an option I could live with. To find balance is to find a passion for what I wanted to do where work is a joy. The reason I wanted to start a company wasn’t because I want to build and run a company (I had done that before), but it was because I wanted to solve the power imbalance.
Our economy is so skewed that the imbalance in power between the monopolies/oligopolies and the rest of us, is vast.
Given my experience in the cloud computing space, there are 3 such providers who form an oligopoly where they own 70+% of the compute spend. Shifting the power balance to the consumers of cloud resources is what aligned my personal motivation with my professional experience.
Talking to the right companies…
I knew my search would take longer this time. It accelerated starting in January 2021.
If I am to work for someone else, I wanted to work in companies that were focused on three areas that I liked. Privacy — what I call *positive* security, Machine Learning, Crypto/Blockchain. I turned down offers and a lot of interviews (something I would not have done before) because it didn’t fit what I was looking for or didn’t align with my definition of meaning and purpose.
I had spent 7 years at CHKP. I had a lot of interview requests for security companies. Most of them, I couldn’t tell what they were offering that was different. Most were, what I call, *negative* security companies.
Positive security is enabling privacy for individuals, making it easy for them to access applications, maintain their passwords, etc. This is a better fit for me, IF, I were to get back into the security space.
I focused on the cloud optimization space. There are a bunch of interesting companies in this space.
- I could join a startup that met one or more of my interests. I could use some of the money to bootstrap it. Would take longer to achieve the goals. There were 2 companies that had the right role for me and fit the Privacy and ML criteria
- Find a company where my purpose aligned with theirs in the optimization space and more importantly, their rationale. There were three companies that fit the optimization and/or ML criteria. Lots of opportunity still to be discovered. This was a higher priority for me.
After interviewing with 5 great companies, I narrowed down to 3 companies that would be the right fit that aligned with my personal mission and purpose.
Option 2 meant, sidelining the building of my own startup.
The key was doing the work that brought meaning and purpose to my daily life if I am to spend a ⅓ of it working for someone else.
As I have been advising other founders and would be entrepreneurs, I ask them why do they want to build a startup? Most talk about solving problems but can’t answer why they care about solving those problems and why them?
For, it’s that internal compass that’s the motivation, so when things get challenging, as they often do, we would have to rely on that motivation to keep us going.
I had this notion of democratizing and decentralizing cloud compute so the balance of power is shifted to the consumers of cloud resources. Building a startup, was a means to this end. If I could do this elsewhere with a great team, so be it!
As Jeff Bezos of Amazon says: “Be stubborn about the vision and flexible about the details”.
What mattered is the mission and vision. How I get there is flexible.
A few words about compensation
At startups, you are not going to get Google or Facebook salaries.
My advice to career professionals is to first figure out what will give your life meaning and purpose and choose a job that fits that. Really understand what kind of life is going to make you happy. Hint: It isn’t about money or glory or power.
Yes, you need to make enough to afford a good quality of life where you live but think hard about what success means, internally, not what society or others think or judge.
My salary requirements are not wealth focused. It’s not cheap to live in San Francisco but the quality of life wouldn’t be better elsewhere for me. There are a lot of intangibles that keep us here. Our friends being at the top of that list.
With my own startup, I would have the largest equity stake. Not doing that, meant that my equity stake will be much smaller and rightly so. I wanted to work on problems and I wanted just enough equity stake + base. I didn’t care about being rich. I wanted to have a good middle class life and have no debts by the time I retire.
To me, salary and equity are independent decisions.
Salary — This is what I am worth today in the market place given my skills and experience. I know that a lot of startups offer equity as compensation in lieu.
Equity — It’s about risk. The younger the startup and bigger the challenge, the riskier it is.
If I had no obligations and I was right out of college with no debt, I would probably think differently. But if someone is hiring me for a leadership role, then the reward potential should match the risk profile. There are no guarantees of course. It’s a gamble, but an educated one.
Crypto tokens as part of equity — This was new. I didn’t know how to evaluate this. I had been studying crypto but never had I received an offer where tokens were part of the compensation. The entire leadership team met with me on a Friday afternoon for 2+ hours to help answer all the questions. They were up front, open, and patient with my naive questions.
I took the weekend to think about it and do some more reading. By Sunday, I was convinced.
I chose Akash Network, the world’s first decentralized open source cloud, and DeCloud for DeFi.
The people there are super smart and their vision aligns so perfectly with what I want to do for the foreseeable future. I could see myself spending the rest of my career helping build that business without regrets. I know I want to give it my best effort and shoot for the moon with them. I am not afraid of taking risks as long as I am not worried about my day to day life.
I am honored and grateful to the team there — Greg, Adam, Boz, Cheng, Maly, and the rest — to have me. I am looking forward to picking their collective brains on crypto and blockchain. Even if they had not hired me, I would still be super interested in learning from Greg and the rest of the team. I can’t wait to be vaccinated so I can see them in 3D :-)
Let’s go build the best cloud compute marketplace and democratize it by decentralizing how providers provide compute so tenants get the right compute, at the right time, for the right price.
Today is day 1 as Head of Product, Akash Network. I’m looking forward to working with the team to help make their vision a reality.