The new skill: Building culture and process to fit a changing world

Most product managers eventually face the same challenge: integrating products, teams, and systems that were never designed to work together.

It happens in partnerships.

It happens in platform integrations.

And in the most extreme version of the problem, it happens during an acquisition.

Nine months ago, we found ourselves on opposite sides of one.

Both of us were newly joined product leaders, pulled into a live M&A process during due diligence, before it was even clear whether the deal would close.

Two companies, two products, two languages, and no shared roadmap, ownership, or even a single source of truth.

Instead of waiting for organizational clarity, we focused on the one thing that could anchor everything: the problem this “product marriage” was meant to solve.

Before the acquisition was signed, we started joint story-mapping sessions and began designing the future product together. That shared view allowed us to navigate uncertainty, align teams across companies, and move forward even while the deal itself was still in motion.

Two months after the acquisition closed, the first customer was already live on the integrated platform.

In this talk, we’ll share what worked, what almost broke, and how product managers can act as cultural translators and integration leaders when different worlds need to become one product.

The new skill: PM as a Builder (Ship-It-Yourself) 

What happens when the person holding the product vision finally has the power to push it directly to production?

Over the last year, as a non-technical VP Product, I bypassed the traditional engineering queue and shipped over 100 features directly to production.

The magic isn't (just) the AI tools; it's the framework we built: Product-Led Development. By deleting the PM-to-dev translation layer, we collapsed the idea-to-production flow into a matter of days.

In this talk, I’ll share the exact blueprint to rewire your organization for this flow: starting with a small task force, defining non-negotiable guardrails, and earning R&D's trust.
I’ll show how this model accelerates business velocity, transforms how you handle customer requests, and ultimately frees your engineering team to focus entirely on core architecture.

The new skill: Product Ownership of Packaging

Product packaging is often treated as pricing or go-to-market decisions. But for customers, packaging is the product; it shapes how they understand what you offer, how they evaluate it, and whether they choose to engaged or walk away.
When packaging reflects internal structure instead of customer value, it creates hidden friction across the entire sales process
In this talk, we’ll examine how packaging shapes customer decisions, where it creates friction across the funnel, and how small decisions can impact conversion and adoption, along with how product managers can help shape these decisions based on a deep understanding of customers and the market.

The new skill: Discovery Without Data

Imagine managing a complex product while navigating through a thick fog with your eyes closed. Due to strict security constraints, your product is installed on-premises with no external access: you are barred from seeing the customers' content domain, you cannot use their data for quality assurance, and you can't even track user actions to understand how they interact with the product on a daily basis.

As a Product Manager at AudioCodes, I faced this exact scenario while working with the Defense and Government sectors. I discovered that standard methodologies, such as analyzing usage metrics or conducting direct user interviews, were simply inaccessible or inapplicable. In this talk, I will share how we cracked the product discovery process in a "closed lab" environment and delivered significant value without ever seeing a single word of the customer's data.

Key Takeaways:

• The Shadow Persona Method: How to characterize needs through organizational "intermediaries," leveraging AI tools to synthesize partial information into a comprehensive user profile.
• Building a Synthetic Golden Data Set: Principles for creating an internal benchmark that accurately simulates customer challenges without relying on sensitive or classified data.
• The Proxy Validation Model: Using indirect signals (installation reports, support tickets, and pre-sales feedback) as quantitative success metrics to validate your roadmap.

In this talk, Dan will break down the five major cost drivers behind LLM calls at scale, along with practical approaches that can reduce costs by 20%–80%.

The new skill: Shipping and scaling AI Agents

Over the past 18 months, we built and scaled an AI-powered code review agent at Baz, now used daily by thousands of developers.

Along the way, we made many architectural, product, and UX decisions that felt reasonable at the time. Some worked, but several turned into costly mistakes that impacted adoption, user trust, and revenue.

In this talk, I share the lessons from building AI agents in production, where there are still very few established best practices. While most talks focus on success stories, this one focuses on what breaks when AI features meet real users.

The new skill: Professional Intuition as a Signal

In a data-driven world, we often silence our inner sensor: professional intuition. In this talk, I’ll share how I navigated leading the product under significant doubt, and how I intentionally stopped in a fast-moving environment to challenge "proven" assumptions. This approach led me to choose a path that ultimately drove a fundamental shift in our results.

AI is changing everything we thought we knew: code is becoming a commodity, enterprise clients expect value that is more tailored every day, and looking different from a generic LLM gets harder with every new model announcement. We had to act, so instead of changing the roadmap again, we changed the product role.

As Head of Product at Tastewise, the AI platform behind the world's biggest food companies, I worked with the founders on a radical move: collapsing four job families into one role, the Product Builder. 8 people (product managers, designers, analysts and solutions engineers) now own problems end to end, sit in every meaningful customer call, and ship to production themselves.

This is the inside story, including the parts that did not work: what broke, what changed, what we quietly built underneath to keep production safe, how the organization and the market reacted, and what actually held.

As product managers, we are expected to move fast while staying constantly aware of everything that can affect the product: competitor moves, customer signals, new models, APIs, agents, platform changes and emerging capabilities.

To make that manageable, I built an AI-based daily intelligence system that continuously fetches from the sources that matter to me, validates what it finds, filters the noise, scores each update by impact, and delivers a short digest with the few things that are actually worth my attention - including why each one matters to my product.

What started as a personal workflow quickly became something other people around me wanted to adapt to their own needs.

I’ll share how I turned product judgment into a reusable Skill + Automation, and how teams can build their own intelligence workflow around the signals they cannot afford to miss.

The new skill: Evaluation as a Daily Tool

When you build products around GenAI, you quickly realize the old rules don’t apply - traditional QA won’t cover you. The product isn’t binary; it’s probabilistic. A tiny tweak in a prompt chain can ripple through the entire experience.

When your product’s core value depends on an LLM’s ability to reason, "hoping it works" isn’t a strategy.

In this talk, I’ll share how we moved from vibe-based decisions to building real confidence in our product. I’ll walk through the messy, iterative process we went through, and how evaluations became a core part of how we build, test, and improve AI-driven experiences.

I'm the only product manager in a growing health tech company, writing requirements for five teams at once: SaaS, embedded, hardware, data, and design. Instead of waiting for headcount, I built myself a team of AI agents - one for each stage of product work, from discovery through user stories and QA to design handoff - working with each other and with me, directly on our real Jira, Confluence, and Figma.I'll show how I built this team step by step, what failed along the way, and why when the stakes are high, the human must stay in the loop.