Why we invested in Griptape: Enabling enterprises to unlock the power of Generative AI

Peterson Ventures co-leads Griptape's pre-Seed round.

Generative AI has the potential to change the way we live and work in ways that are still hard to imagine. At Peterson Ventures, we see the introduction of large language models (LLMs) as a paradigm shift in software development akin to prior technology shifts with mobile and cloud computing. In order to build software leveraging LLMs, a new technology stack is needed. 

Generative AI x Future of Work

At PV, we invest at the earliest stages, typically writing our first check into companies at the pre-Seed and Seed stages. Right now, nearly every new startup that pitches us is trying to solve a variety of pain points in novel ways that include the use of LLMs. This is especially true in our Future of Work practice, where we are seeing founders building a variety of innovative solutions to automate rote tasks, painful workflows and frustrating bottlenecks leveraging LLMs. The way knowledge workers interact with their day-to-day tasks is going to completely change. 

Based on the recent flood of pitches we’ve heard from founders leveraging generative AI technology and the speed at which incumbents like Microsoft Office, Google Workspace, Adobe, Canva, Notion, etc have incorporated Generative AI features into their products, it quickly became abundantly clear that most software companies, large and small, will start incorporating Generative AI features into their applications.  But a lingering question we’ve had as a team is: what are the middleware tools needed for  a mid-market or enterprise company to take advantage of the power of LLMs while still protecting their proprietary data, personal identifiable information (PII) and other sensitive data? Generative AI spend is forecasted to grow 35% YoY reaching $109B by 2030. In order to unlock a good portion of this spend, we believe the question of how enterprises can securely use LLMs needs to be solved.

The Technology

With this question top of mind, we were prepared to move quickly when we met Kyle Roche and Vasily Vasinov to learn about what they are building at Griptape. Kyle and Vasily mirrored back to us the question “how can enterprises take advantage of LLMs safely and securely?” Their answer was Griptape, an open source app-building framework purpose built for enterprise. Griptape enables mid-market and enterprise customers to safely bring LLMs into their organizations to build complex, secure applications, including conversational, copilot and autonomous agents. The framework gives enterprises security controls around data sources and ways to interact with LLMs so that they can leverage the power of LLMs to solve complex workflows without compromising data security. As we dug into the product and technology, we were intrigued by the novel approach Kyle and Vasily were taking to solve this problem.

Source: https://github.com/griptape-ai/griptape

As a metaphor, think of remodeling a house. Today, leveraging a LLM to build an enterprise application is akin to opening up the front door of your house wide open, and giving every sub contractor total access to the whole house – the painter can go check out the plumbing, the plumber can wander over to open up the fridge, and anyone can go through the closets. Creepy, right? Which is exactly why an enterprise wants to limit data access and control privacy when building applications using generative AI models.  

In contrast, Griptape’s technology is like hiring an excellent General Contractor, who carefully controls access to each room of the house. The plumber only has access to the pipes. The painter only sees the wall that’s getting painted right now. There’s no wandering access over the whole house. 

While the landscape is moving super fast and it’s not yet clear what types of companies will win the application layer as “sub-contractors” (horizontal, vertical, incumbent, etc) we know that having this strong “General Contractor” middleware layer is going to be necessary before enterprise companies can truly use Generative AI and LLMs with confidence. Griptape is the solution that unlocks this for all applications.

The Team

Our conviction grew as we got to know Kyle and Vasily, the co-founders of Griptape. Something we look for when backing founders is if they have a unique insight based on their lived experience into the problem they’re solving - these two founders demonstrated this in spades as we dug in. With their unique combination of early stage startup and enterprise experience, Kyle and Vasily demonstrated a knowledge of what it takes to build an enterprise product while operating at early stage startup speed. 

Kyle, co-founder and CEO, was previously co-founder of 2lemetry, which was acquired by Amazon. He went on to work as a GM at AWS for 8 years and built more than 10 public services during his tenure there. Vasily and Kyle met when Vasily joined 2lemetry as a senior software engineer, and he continued on with Kyle at AWS, where he worked both as a software engineer and a principal product manager. The two founders have worked together on and off for over a decade, are deeply aligned with a shared vision for the product and technology they’re building and equally understand what it takes to build an enterprise grade product and a startup from scratch—they were a team we had to back.

Announcing Griptape

We are delighted to announce that Griptape came out of stealth today with the announcement that Griptape has been selected to be in the inaugural cohort of the AWS Generative AI Accelerator (check out the announcement by AWS) and that Range Ventures and Peterson Ventures co-led their pre-Seed. A huge thank you to our co-investor Chris Erickson, co-founder and Managing Partner at Range, for introducing us to Kyle and Vasily—we couldn’t imagine better partners to work alongside as we support the Griptape team building into their vision. If you’re looking to using Generative AI at your mid-market or enterprise company, check out Griptape on Github.