The Zetta AI Founders’ Playbook
Lessons from the most successful AI-native companies
At Zetta Venture Partners, it's our mission to help technical founders turn machine learning models into market-leading companies. We were the first VC firm exclusively focused on identifying and supporting AI-driven, B2B businesses — and we remain as confident in this thesis as we did when we began, given the growing transformational impact of applied AI in critical domains like biotechnology, climate technology, DevOps and cloud infrastructure, security, and more.
Since the launch of our first fund in 2015, we’ve seen the commercial AI ecosystem transform: models and data have become more sophisticated as compute has become more accessible, and AI and ML techniques have come to be seen as a competitive edge with clear value delivery. Over the last several years, increasing numbers of AI-native companies have emerged and been successful, helping businesses solve problems in dozens of sectors.
We’re excited to introduce our latest project: the Zetta AI Founder’s Playbook – a resource for applied AI companies based on real lessons learned by breakout AI-native startups and veteran industry operators. As investors, operators, product leaders, and engineers ourselves, we’ve had a front-row seat to enterprise AI company-building in a wide array of contexts. We’ve met, partnered with, and learned from the leading founders and researchers who paved the way and built the AI-driven products that created the field as we know it today–and think that there are important lessons to spotlight about the unique aspects of building a business in this technology area.
We think that AI companies are different than traditional software companies in some important ways:
- AI is hard to build — it requires different skills and expertise, and bilingual founders that understand the domain and data as well as ML modeling and infrastructure.
- AI doesn’t stand alone — it is dependent on upstream inputs (data, labels, integrations) and downstream outputs (workflows, feedback loops).
- AI takes time to start working — it usually requires upfront investment by early users, creating some which-comes-first (cold start) and other strategic challenges that traditional software companies don’t face.
- AI performance is not consistent or guaranteed — it can get better by learning overtime, but it can also degrade (drift) and malfunction (edge cases). Performance is data dependent and may vary across customers.
- AI businesses can take many form factors — subscription software is one of many, and, as we’re finding, may not be a good fit for many applications. There are nearly as many business models as there are businesses in emerging AI markets.
- AI is high value, high risk — systems that make decisions and change over time can be higher value but higher risk than those that manage workflows or execute static operations.
- AI presents unique sensitivities — all businesses should be deliberate about product development and deployment, and AI use-cases deserve special attention to ensure that they do not cause unintended consequences, bias, or misuse.
Over the coming months, we’ll be interviewing founders and operators spanning the industry, with the goal of dissecting and disseminating strategies, tactics, and stories relevant to AI-driven businesses. Check out our first posts below, and subscribe to our newsletter to follow along as we build the Zetta AI Founder’s Playbook, one applied AI success story at a time.
Strategy + Tactics
- The AI-first startup playbook
- AI adoption is limited by incurred risk, not potential benefit
- Data rights are the new IP rights
- The Intelligent Enterprise Stack
- Beating Behemoths
- Framework to grasp industrial analytics opportunities
- Positioning An AI-Native Company
- The Intelligence Era and The Virtuous Loop
- Zetta Bytes AMA: Hiring a CTO
- AI Entrepreneurs: 3 things to check before you pursue a customer
- There are more data scientists than you think
- Stages of funding in the intelligence era
- Zetta Bytes AMA: Questions to ask about pricing
- Could data costs kill your AI startup?
- Measuring AI startups by the right yardstick
- Finding the Goldilocks zone for applied AI
- Data is not the new oil
- How Synthetic Data is Accelerating Computer Vision
- Machine Learning in the Deployment Age
- Innovations of the Next Decade
- Computing Like a Human