AMA: PayPal Senior Director of Product Management AI Platform, Jarred Keneally on AI Product Management

I look for people who use AI, not just talk about it. I look for: Concrete examples: Show me your n8n workflows, your automation stack, how you've 10x'd output without sacrificing quality. How you connect github, jira, and confluence with MCP and redefine how you do product planning. Creative application: Beyond the obvious—research, competitive intel—how are you using AI for scenario planning, GTM experimentation, or stakeholder alignment? Growth mindset: Candidates who experiment, iterate, and ...Read More

AI is shifting product marketing from production to orchestration. The bulk of my time used to be spent on execution—drafting positioning, building decks, writing briefs. Now it's about defining strategy, curating inputs, and refining outputs.What I can automate: First drafts: Messaging frameworks, competitive analysis, launch briefs. Tools like n8n pipe research into templates that get me 70% there. Content variants: A/B test copy, regional adaptations, persona-specific messaging—at scale. Data ...Read More

At first yes, but at PayPal we have moved beyond just basic productivity improvements to a company wide initiative becoming AI Native. identifying opportunities across all company functions to eliminate silos and develop strategic AI solutions to bolster the achievement of company goals.​The limitation isn't the technology—it's our deeply ingrained beliefs about who can do what.​Our roadmap looks like this: BU Journey Mapping >> Strategic Evaluation >> Identifying the Native AI oppor ...Read More

The skills that matter aren't new—they're amplified. AI doesn't change what you need to be good at; it changes how much those fundamentals matter.What becomes critical: Prompt engineering as a core skill: Your ability to extract value from AI tools directly impacts output quality. This isn't optional anymore. Data literacy: Understanding what signals matter, how to structure inputs, and how to interpret AI-generated insights. Strategic judgment: AI accelerates execution, which means bad strategy ...Read More

Influence comes from showing up with data, not opinionsHow I build credibility: Bring customer data: Usage patterns, support tickets, feedback, analysis. "Customers are asking for X" backed by actual quotes and frequency. Competitive intelligence with context: Not just "competitor shipped Y," but "here's why it matters and what customers are saying about it." Business case, not feature requests: Tie AI opportunities to revenue, retention, or efficiency metrics Product already cares about. Protot ...Read More

Everyone's claiming "AI-powered" everything. Cut through it by anchoring on two things: defensibility and outcomes.What is your core value? Generic LLM wrappers are commodities—your competitive edge is what only you can train on or optimize for.Don't build AI because competitors are. Build it where it eliminates a painful workflow or unlocks something previously impossible.Features that get better with usage, scale, or network effects.Focus on where AI creates measurable customer outcomes tied t ...Read More

This is a great question. Personalization at scale is where AI delivers real ROI. Here's a concrete example from my workflow:I analyzed the tone, style, and grammar of emails from myself, our CTO, and my manager—fed that data into prompts to generate content in their distinct voices. Now I can draft executive communications, technical updates, or stakeholder product updates that sound authentic without starting from scratch every time.Claude Skills are great and I have defined the Claude Brand S ...Read More

While I cannot share a lot on our use cases, Voice is a critical function across all agents for agentic commerce. Multimodal Integration touch, speech, text, images, and video is needed.Other Emerging Trends Cross-channel reconciliation agents: Automatically detecting and resolving discrepancies across web, mobile, POS, and invoicing—no manual intervention. Dynamic payment routing: Agents optimizing for success rates, fees, and compliance in real-time based on customer context, geography, and tr ...Read More

Yes, we use backstage and a home grown developer portal. Where backstage excels is in creating one place for all of your tools, documentation, services and observability. We have created agentic ops, that is a single agent that walks a user through the CI/CD pipeline for managing, deploying and monitoring. This works well because you can truly enable more users to effectively deploy solutions, but all the guard rails for security, source control, and compliance are enforced with out causing the ...Read More