Snyk Innovation, Technology & Agility

How does your team use AI in the engineering design process, and what benefits have you observed? How does Snyk support and encourage you to use AI safely?

My team works with our product management and engineering teams to design new features, improve existing ones and write technical documentation for Snyk products. 

Snyk encourages the safe and effective use of AI. We have a dedicated AI governance team that guides us on risks and opportunities. We use AI to make our workflows faster and more efficient. During brainstorming sessions, AI helps us organize and evaluate ideas. It also acts as a creative partner during ideation, suggesting alternative solutions that keep us focused on customer needs. 

On the documentation side, we are piloting AI to generate the first draft. We want to better understand if we can accelerate creation of high-quality and accurate documentation of all Snyk product features. 

Finally, we’re using it for some of the most impactful solutions, such as fixing security. 

Snyk’s balanced approach to AI, which encourages engagement and experimentation with the technology while maintaining appropriate governance and education, is key to its continued adoption in the company.

What challenges have you encountered when implementing AI technologies in engineering design, and how have you addressed these issues?

Like any new technology, AI brings excitement and uncertainty. On my team, there are varying levels of fluency with AI tools. Externally, the hype around AI often exaggerates what it can achieve. 

One challenge we faced was identifying meaningful use cases and trusting AI-generated results. For example, AI can hallucinate, leading to inaccurate conclusions about user needs. To address this, we leveraged our internal large language model for insights on select interview data that had already been manually analyzed. The LLM captured broad themes accurately but fabricated some customer quotes. To fix this, we refined the prompts and now use the LLM only for identifying high-level themes. Qualified team members still review and validate all insights. 

Another challenge was using AI for design solutions. While generative AI can produce numerous iterations from a single prompt, it often misses the mark on creating specific, usable designs. For example, I couldn’t arrive at the intended design idea solely through AI prompts. We’ve learned that AI works best for freeform brainstorming, but designers still refine these ideas into solutions that align with the product’s look and behavior. AI is a valuable tool, but human expertise remains essential to ensure quality and alignment. 

What excites you most about your team’s future when it comes to leveraging AI in innovative ways?

AI’s ability to retain context and process vast data across domains allows us to make faster, more informed decisions. As the technology evolves, I see it enabling a more human-centric and efficient way to interact with systems. For instance, instead of writing complex queries to understand security vulnerabilities, customers could simply use prompts to get faster, more precise results. 

AI will transform our internal workflows. It can detect bottlenecks early on and optimize team processes. It can ensure that design and technical documentation is accurate, high-quality and compliant with organizational standards. By flagging potential communication gaps or problems early on, AI will help us address challenges proactively. 

With proper governance, AI can enhance efficiency while safeguarding against risks. I see AI as a partner, not a replacement, for our teams. I am excited to see how empowered teams will leverage AI responsibly to accelerate innovation at Snyk and across the industry.