Editor’s note: Sudip Chakrabarti is a Partner at Decibel, where he works with companies that are transforming the enterprise, cloud, and infrastructure markets. Prior to becoming an investor, he was an entrepreneur and an engineer, starting his first company when he was at Georgia Tech, and his second while in business school. Sudip earned his Ph.D. in computer engineering from Georgia Tech and his MBA from The Wharton School of the University of Pennsylvania.

Can you tell us about your background and how you came to Decibel?

I trained as a computer science engineer at Georgia Tech before becoming an entrepreneur and starting two software companies, both of which were successfully acquired. Then about 10 years ago, I decided to get into venture investing. I was a partner at Andreessen Horowitz and then at Lightspeed and Madrona. As an investor, my focus has always been exclusively on enterprise infrastructure, and I’ve been fortunate to back what have gone on to become some pretty big companies, including Heptio, Exabeam, Spotnana, Menlo Security, and Temporal.

A few years ago, I met Jon Sakoda, Decibel’s Founding Partner. Having previously been a technical founder myself, I really liked his vision of working with technical founders at the earliest stages to provide highly specialized help. That’s pretty rare these days considering that most venture funds have gone multi-stage and multi-sector. About a year and a half ago, I joined Decibel to try to help them realize that vision and build the kind of VC firm that I wish I’d had back when I was a founder myself. 

What is Decibel’s investment thesis? 

We invest exclusively in early stage B2B infrastructure software companies, meaning that we partner with technical founders who are building technical products that are sold to technical buyers. In other words, products used by developers, engineers, data scientists, and cybersecurity and IT teams, among others. 

Within B2B infrastructure, we largely focus on companies that have a bottom-up go-to-market strategy, whether that’s via open source or other means of grass roots adoption such as freemium models. In terms of stage, we strictly focus on Seed and Series A, but also reserve considerable funds for follow-on rounds. 

And beyond that focus, what differentiates Decibel from other VCs?

Our differentiation starts with our specialization. Each of us at Decibel has only ever operated and invested in the B2B enterprise infrastructure space. As a result, from our very first engagement with a founder, we’re able to connect on a deeper level and provide a very differentiated and high-fidelity experience. And while our expertise manifests in everything that we do, it’s particularly relevant to how we help founders post investment. 

What does that post-investment help typically look like?

We provide a suite of founder services that technical founders look for at the early stages of building their companies. We spend a lot of time helping our founders craft and refine their go-to-market strategy, helping them with product-led growth, community adoption, and product positioning, as well as finding them early customers. We have a deep and broad network of tens of thousands of early adopters and customers who we can connect our founders to. They’re director and VP-level professionals who actually have the pain points our founders are working to address, as well as the budget and decision-making authority to take action. 

Beyond that, we help with recruiting. I’m not talking about C-level executives, but rather the first 10 to 25 team members, which for our founders typically means engineers and product managers. We also have partnerships with PR firms to help our portfolio companies get publicity around their product launches and brand building efforts, and with pitch coaches to help them craft their narratives.

If you take any one of these services on its own, it’s probably not all that unique. But when you offer all of them together and they’re all tailored to one very specific sector and stage, it becomes highly differentiated and compelling.

Decibel invests in a lot of open source companies. Why is that and can you tell us about any recent investments?

Within B2B infrastructure, there have been a couple of different go-to-market motions. The first was sales led, which was a very top-down approach where sales people would go out and sell to their customers. Then came product-led sales, where the goal was to bring people into a product to trigger a sale. We’re now in the third iteration, which we call the community-led motion. At today’s most successful infrastructure software companies, sales happen when there’s a massive user community behind the product that gets value out of it before some of them eventually convert into paying customers.

quote icon

At today’s most successful infrastructure software companies, sales happen when there’s a massive user community behind the product that gets value out of it before some of them eventually convert into paying customers.

Sudip Chakrabarti

In our world, open source is hugely important because it helps drive community-led sales, so we try to marry that understanding with our investment theses. It was with that in mind that we recently invested in a company called Penpot that we believe is poised to lead what we’re calling the DesignOps movement. Similar to how DevOps has been so transformational over the past 15 years in terms of bringing much greater efficiency to how software development and IT operations teams work together, we see an opportunity to rethink how designers and developers interact. 

The problem is that today, most designers use whatever tools they want before passing their designs on to developers. As a result, iteration is often a slow and clunky process. Penpot solves that problem by not only being a really great design tool that’s attractive to designers, but also open source, which developers like because it allows them to build integrations to their tool chains. We really liked the team at Penpot and their vision, and quickly realized that it was the best solution out there to make DesignOps possible. In September 2022, we led their Series A. 

Based on your work, what advice do you have for founders trying to build successful open source companies?

Three things come to mind. First, to really succeed as an open source company, you have to make your open source community successful. Balancing the needs of the business and the community isn’t easy, but if you get it right, are authentic, and genuinely care about the community, you’ll have multiple ways to monetize it down the road as you build your business. 

Second, don’t rush into monetization at the expense of adoption. If you can really nail adoption and become the standard for the space you’re in, you’ll have a lot of opportunities to monetize over time. The reality is that successful open source companies have a longer gestation period than most other businesses. But once they take off, they’re able to scale much more quickly because they already have their top of funnel in place.

And what’s the third thing?

Even though you might not monetize for years, you need to know what your monetization plan will be from day one. Doing so will help you decide which parts of the product you put out in open source and which parts you hold back. If you give away too much, you’re basically giving away the company. Don’t give away enough and you risk starving adoption. Our rule of thumb is to give away everything that makes individual users successful, while holding back things that enable success at the team and enterprise level. 

Another area where you spend time is large language models. What trends are you seeing there and can you tell us about any investments you’ve made? 

If you look back over the past 20 years, there’s been a platform shift every decade. In the early 2000s it was mobile. In the early 2010s it was cloud computing. Right now, large language models feel like that shift for the 2020s. And while there’s certainly a lot of hype around it, I genuinely feel that we’ve reached an inflection point. The technology has finally reached the stage where in spite of lacking reasoning abilities, it can do amazing stuff in a large number of use cases.

In terms of where we are focusing, if you think of a tech stack, there’s the infrastructure layer at the bottom of which are the public clouds. Then comes the foundation model providers like OpenAI, and then the enabling tools and frameworks on top of that. We’re spending a lot of time in that top layer. More specifically, we’re focused on LLM native applications and how founders are applying LLM technology to existing categories to reinvent the user experience and cost structure, as well as latency and performance. 

We’ve made several investments in this area so far. One is in stealth mode and helps developers write secure code, using LLM and generative AI technology. Another investment is called Botpress and is building customer support tools like chatbots using LLM technologies. We recently made four other new investments in the generative AI space that are all in stealth and look forward to announcing them in the coming months.

Given your knowledge of LLM, what’s your take on ChatGPT? Is all of the hype well founded?

quote icon

What ChatGPT did was make the [LLM] technology accessible to the masses.

Sudip Chakrabarti

From an investment standpoint, we’re not particularly swayed by ChatGPT. Our companies, for the most part, were already using the technology you’d find under the hood of ChatGPT. The GPT-3 model was released back in June 2020 and GPT-3.5 was released in March 2022, so they’ve been around for a while. 

What ChatGPT did was make the technology accessible to the masses. My wife, for example, works in healthcare and can go to ChatGPT and ask for a treatment plan for a patient with a particular disease. The output that comes back may not be perfect, but it shows her and anyone else using ChatGPT what’s possible with generative AI. I think that’s the biggest benefit of ChatGPT. 

With all of the disruption and uncertainty over the past few years, what are your thoughts on the future of venture capital? How should early stage VCs and founders be thinking differently about the industry? 

As early stage investors, we take a long-term view. Most of our companies won’t come to market for a year or two, before scaling and eventually catching the next growth cycle. So we have the luxury of time and don’t have to have the knee jerk reactions you might see if we were sitting on a portfolio of late-stage companies that have already raised a lot of money, are burning tons of cash, and are way ahead of their valuations.

From our perspective, it’s a great time to build a company. Talent is more available, customers are more open to trying things that don’t cost a lot because they’re working with tighter budgets, and those same customers are generally more willing to give vendors their attention in this kind of environment.

I’d also add that the more things change, the more they stay the same. Venture capital really has its roots in being a trusted advisor, helping founders avoid the mistakes they made themselves or that they saw others make. Over the last few decades, however, as the industry has scaled and gone multi-stage, it has broken into two groups. Company builder venture funds help founders build their business from the ground up, while company scaler venture funds help more established businesses accelerate their growth. 

I think thoughtful founders have realized two things. One is that capital doesn’t solve many early stage growing pains. What they really need in those early days is impactful help — help that moves the needle and greatly increases their odds of success. The other thing founders have realized is that you can build a lot in terms of a product before needing to consume a lot of capital. I think for those reasons old fashioned early stage venture investing is back in vogue. Beyond just writing a check, we give founders the help they need early on to build their business.

Thanks, Sudip! We appreciate your insights.