Bringing Generative AI into Product: Shay Shitrit of Augury
Host: Eyal David · Guest: Shay Shitrit · About the host
In this episode
- AI at the discovery stage makes it possible to distill insights from an enormous volume of sources (transcripts from Gong, Slack conversations, CSM/RSM summaries, presentations) within hours instead of weeks — Shay says he won't do discovery anymore without this tool.
- Claude's large context is what made it possible to feed in all the raw material at once; the preparation work is 'Sisyphean' (3–4 hours of collection) but the output justifies it.
- The product skills he considers most important: high EQ and empathy, leadership without authority (alignment and influence over other squads), and a just-do-it, execution-focused approach that avoids analysis paralysis.
- Adopting GenAI tools in an organization doesn't happen on its own — you need to spread the word, success stories, a sense of FOMO and accessibility in a single link; discovery has higher friction than the spec generator and therefore lower usage.
- The end users aren't the classic technological SaaS users — technicians on a production floor who may have no computer, no smartphone or no internet — which requires solutions that cross into the physical world (SMS, email, even a TV in the break room).
- Augury's strategy leans on the enormous historical data about the customer's machines as a differentiation lever; the GenAI in the product (talking to the machines) is trained on this data and creates a value loop that comes back inward.
Listen to this episode
Episode Description
Shay Shitrit is a product manager at Augury, a machine-health company that monitors machines with sensors and prevents unexpected failures on the production floor. In the episode he tells the story of his journey from industrial engineering and management at the Technion, through Unilever and a failed startup, to Augury — and about what he's leading today: bringing Generative AI tools into the work of the product department. The conversation dives into how he and his partner Dov built AI-based discovery processes, a spec generator and internal agents, and how you embed such technology in an organization when the end users are technicians on a production floor.
Full transcript with speakers and timestamps
Selected quotes
“I'm probably not going to do any discovery without using this tool.”
“Do, break, learn, improve. Go out fast, learn, and fly off. As long as the general azimuth is good, it'll be fine.”
“I leave it to future Shay. Future Shay is better, he's smarter, he has more experience, he has a different perspective.”
Questions & answers
What does Augury do?
Augury is a machine-health company that, using sensors and physical installations, monitors and samples machines in production every hour, identifies developing failures months in advance and recommends what to do — thereby preventing Unplanned Downtime instead of the wasteful route-based method (a technician passing through with a vibration meter once a quarter).
In which product areas is the GenAI applied?
They mapped the areas a product manager touches — discovery, spec, release notes and more — and chose to focus on two: discovery (distilling insights and surfacing problems from all existing information) and the spec generator (an agent that builds an action plan, defines a problem and a solution, and what to measure).
How did the AI-based discovery process actually work?
They gathered many sources — transcripts from customer calls on Gong, context from Slack, CSM/RSM call summaries, presentations and docs — fed them into Claude thanks to its large context, and distilled insights. The next stage was validation with internal stakeholders, and the result entered as a meaningful backlog that connects to the roadmap.
How do you embed a GenAI tool in an organization when there's no precedent?
Shay and Dov started with spreading the word — showing the output, sharing success stories in the PM channels and the product guild, generating a sense of FOMO ('an hour to make a spec like this?') and making the agent accessible via a single link for free use (totally vibe). Without an initial push and help with adoption it simply doesn't happen.
What are the three product skills he considers most important?
High EQ and emotional intelligence (empathy and forming connections), leadership out of understanding rather than authority (alignment and influence over other squads as a domain expert), and a just-do-it approach — a focus on execution, moving fast and avoiding analysis paralysis.
Related Episodes
How to Build AI at the Edge of the Device — Ori Katz, Hailo
00:40:49 · March 5, 2026
Ori Katz, who leads the AI SoC domain at Hailo, tells the story of how an Israeli chip company became a serious player in AI at the edge — smart camer...
Amit Slutzky | Trullion | “I Left Figma and Went Into the Field — It Changed Everything”
00:41:22 · February 24, 2026
Amit Slutzky, Director of Product at Trullion, on one of the most important muscles for product people: understanding what's really going on with the...
Tom Mayblum | MIND | When Is a Customer Conversation Worth More Than Data?
00:33:39 · February 9, 2026
Tom Mayblum, VP Product at MIND, in a short tachles episode on when a direct customer conversation is worth more than looking at the data — and when y...
Keren Wexler | Optimove | Product-Team Processes That Create Big Impact
00:44:13 · December 23, 2025
Keren Wexler, Senior Director of Product at Optimove, on product, teams and decision-making — and the point where process stops being bureaucracy and...