T-Mobile

- Leading teams to deliver contextual experiences through AI for T-Mobile customers via digital channels.
- Defining mission, vision, and OKRs for the contextual AI team, aligning with organizational goals and priorities

Principal Product Manager - Technical AI/ML

To The SHE Power

Our mission is to provide women with the tools, inspiration, and connection to boldly design a life that surpasses their every aspiration. We offer mentorship, coaching, and career services to women in transition.

Member of the Board of Advisors

Product School

• Share industry insights through webinars to help inspire the next and current generation of Product Management. 

Product School is a global leader in Product Management training with a community of over one million product professionals.

Featured Speaker at Product School

Instructor of the 'Product Management Course' (3 cohorts) and 'Senior Product Manager Certification' (1 cohort - September '22), online

An 8-week part-time course to build product management skills, covering topics such as product strategy, product discovery, product delivery, and go-to-market.

Instructor

Nike

- Orchestrated multiple product workstreams across engineering teams, driving cross-functional execution
- Directed AI-driven product discovery, optimizing ROI and aligning with strategic business objectives
- Achieved a $100M annual reduction in product returns through innovative AI solutions
- Eliminated apparel testing timelines by implementing predictive analytics models
- Scaled ML engineering team and implemented agile methodologies, boosting productivity

Lead Product Manager - ML & Innovation

The Walt Disney Company

- Developed comprehensive product strategy for the Enterprise IAM team
- Prioritized critical IAM products and features in the engineering backlog

Product Manager, Identity and Access Management

4th Rotation - Product Manager, Machine Learning & Computer Vision Platform: Led machine learning Cyber Security PoC, achieving high accuracy in attack prediction 

3rd Rotation - Product Manager, Machine Learning Platform: Drove AI/ML initiatives for content recommendation systems and personalization

2nd Rotation - Product Manager, Internal Business Onboarding Platform: Spearheaded 'Onboarding at Scale' initiative, significantly reducing team dependencies

1st Rotation - Change Manager, Financial Platform: Implemented KPIs and metrics for adoption tracking, informing executive decision-making

Product Manager, Technology Management Rotation Program

LaunchBoard.io

Technology Management Consultant

WeSolv

- Led product launches, aligning cross-functional teams in an Agile environment
- Engineered device lab, improving build accuracy for mobile applications
- Organized Hackathons and TechTalks, driving innovation and talent acquisition
- Managed and mentored a large team of employees for key initiatives

Technical Product Manager

AT&T

Senior Program Manager, Product Realization

Samsung

Vendor/Device Manager

As a forward-thinking AI/ML Product Leader, I've driven innovation in contextual AI, advanced analytics, and machine learning across multiple Fortune 500 companies. My passion lies in leveraging cutting-edge AI technologies to solve complex business challenges and enhance user experiences.

Throughout my career, I've led cross-functional teams to deliver game-changing AI solutions that have significantly impacted bottom lines - from reducing product returns by millions to revolutionizing personalization in content delivery. I thrive on translating intricate AI concepts into strategic product roadmaps that align with overarching business goals.

My expertise spans the entire AI product lifecycle - from ideation and discovery to development and go-to-market strategies. I'm adept at collaborating with data scientists, engineers, and business stakeholders to create AI-driven products that meet and exceed user expectations.

My commitment extends beyond product development to demystifying AI for product teams and executives. I firmly believe that shared understanding is key to successful innovation. To this end, I regularly lead workshops and mentoring sessions to bridge the gap between technical capabilities and business applications.

As AI continues to reshape industries, I'm thrilled about its endless possibilities. I'm always eager to connect with fellow innovators and leaders who share a vision for harnessing AI to create meaningful impact and drive business growth.

Let's connect and explore how we can push the boundaries of what's possible with AI in product management!

So many things are happening daily that it’s nearly impossible for PMs to keep up with AI.

It’s another paper, repo, blog, or regulation update every week. Meanwhile, leadership wants you to explain RAG, vector databases, or the EU AI Act right now.

Most PMs end up:
▪️Bookmarking endlessly, but never catching up.
▪️Switching tools that forget context or hallucinate.
▪️Running out of time before the next big question lands on their desk.

That’s why I’ve been using NotebookLM. It’s not just another AI assistant:
▪️It only works with your sources (no hallucinations).
▪️Handles far more context than most tools.
▪️Turns research into briefing docs, FAQs, timelines—even podcasts you can listen to on the go.

I used NotebookLM to deepen my understanding of AI Drift—a topic most PMs struggle to explain clearly.
It turned dense research and vendor docs into structured insights, ready-to-use definitions, and a podcast-style recap I could reference anytime.

As PMs, our edge isn’t just “knowing AI.” It’s learning it fast enough to stay ahead.

What AI concept will you master 10x faster next?

Full breakdown here:

Shaili Guru

Feed | LinkedIn

JSESSIONID

li_at

linkedin.com

aipmguru.substack.com

How Product Managers Can Learn AI 10x Faster with NotebookLM

AI/ML Product Leader | Driving Innovation with AI | PM Mentor and Advisor

Drake Karelas: 200 applications, 0 offers.

Then he built an AI agent that changed everything.

Tracked every rejection. Ran self-checks every 50 apps. Built 4 resume templates. Banned AI buzzwords like "revolutionize" and "demystify."

1.7% → 7% callback rate.
3 offers in one week.

The twist? He's sharing the entire system, including his master prompt, completely free.

Why? Because he was drowning in rejections 3 months ago and knows what it feels like.

AI didn't do the work. It made the work smarter.

Full breakdown + master prompt:

Open article: Real Talk AI: How Drake Used an AI Agent to Land His Next PM Role by aipmguru.substack.com

Real Talk AI: How Drake Used an AI Agent to Land His Next PM Role

Every time I teach this course, I’m reminded how fast the field of AI product management is evolving and how much curiosity and collaboration can accelerate learning.

Excited to be co-teaching the AI Product Management Specialization at the University of Washington this fall with Dan Sears!

Can’t wait to see what our Autumn 2025 cohort builds next.

AI Chief Product Officer || Leadership Coach || Strategic Advisor || AI Teacher || ex: MSFT EBAY

I’m excited to share that, together with Shaili Guru, I’m co-teaching the Specialization for AI Product Management at University of Washington!

We've designed this course to be about doing, not just learning. Informed from real-world AI and ML product projects at companies like Nike, T-Mobile, Microsoft, Meta, Xerox Parc, etc. It’s practice-based and exercise-heavy. 42 hours spread over 14 weeks that goes beyond gaining "AI speak". Students form small teams and deliver a significant capstone project.

The curriculum:
-Foundations: AI PM Role & Foundations of AI
-Foundations: Core Concepts, Frameworks, ML Models
-Foundations: AI Product Lifecycle, MLOps & AIOps
-GenAI: Prompting, Context Engineering & Vibe Coding
-GenAI: Experimentation, Iteration & Prototyping
-GenAI: Model Evaluation, Methods & Measures
-Responsible AI: Privacy, Ethics & Governance
-AI & Data Infrastructure: Systems & Infrastructure
-AI Agents: Automation, Workflows & Tools
-AI Agents: Building/Hands-On Exercises
-AI in Practice: Operationalizing AI Across Product
-AI in Practice: AI GTM & Adoption

Welcome to our Autumn 2025 Cohort!
Brietta Easterlin, Angela Cirulli, Lance Traicoff, Lauren Witt, Neeharika Sinha, PhD, Alex Kim, Justin Russell, Justin Mills, CFA, Randy Covert, Wyatt Dobbs, Christopher Fiorelli, Syed H., Nathan Marroquin, Christina Nofziger, CSM, Mike Palumbo, Mithula R., Llew Roberts, Zade Safadi, MBA, Lalitha Tarini, Kari Fausey, Piyu Roy

He had every certification - PMP, Scrum, Agile.

His question: “Do I need another one to break into AI product management?”

It’s a fair question. Certifications give structure and signal credibility.
But they aren’t the differentiator in today’s market.

From teaching AI Product Management, I see a clear pattern:
▪️The PMs who stand out don’t just collect certificates
▪️They get hands-on with AI tools
▪️They learn how data, models, and guardrails actually work
▪️And most importantly, they connect AI to real business outcomes

Certifications can open a door.
But skills build trust once you’re inside the room.

There’s no better way to learn AI than to build with it.

Next week, Hemant and I are hosting something a little different, a hands-on build session for beginners in Bellevue.

🧩 Vibe Issue: Getting Started
 📅 Wednesday, Oct 15 · 11:00 AM – 1:00 PM
 📍 Bellevue Library

We’ll help you set up your environment (Claude Code, Replit, or your tool of choice), and by the end, you’ll have something real, a working prototype, a small app, or even just your first “hello world” with AI.

This isn’t a talk or a panel. It’s two hours of doing - the kind of practical, slightly chaotic, deeply satisfying learning that every Full-Stack AI PM needs to experience.

If you’ve been saying, “I want to build something with AI but don’t know where to start,” this is your sign.

luma.com

Vibe Issue - Getting Started · Luma

Carnegie Mellon | Product | SaaS | User-Centered AI

PMs aren’t being replaced.
We’re being replicated.

One of me in the stakeholder meeting.
One of me using AI to explore edge cases.
One of me drafting the launch doc in parallel.

Same brain, just multiplied.

Last week, I walked out of a meeting to find a draft doc waiting for me.
Not because someone else wrote it because AI had already shaped my notes into a first draft while I was presenting.

That’s not replacement.
That’s amplification.

The real question isn’t “Will AI take my role?”
It’s “How do I scale my impact when there are multiple versions of me at work?”

What’s the smartest way you’ve used AI to replicate, not replace, yourself?

Anthropic quietly launched something last week that could change how we test models: Petri, a framework that automates behavioral auditing.

And no, it isn’t a Petri dish like the one from biology class—though the metaphor fits.
It’s a lab of sorts, but for model behavior.

Instead of checking whether a model can do a task, Petri looks at how it behaves while doing it.
▪️Does it cooperate with misuse?
▪️Does it lie to protect itself?
▪️Does it notice when it’s being evaluated?

When Anthropic ran Petri across 14 major models and 111 scenarios, a few patterns stood out:
▪️Claude Sonnet 4.5 and GPT-5 showed the strongest safety profiles.
▪️Gemini 2.5 Pro, Grok-4, and Kimi K2 were more likely to deceive or bend rules.
▪️Every model showed at least some questionable behavior.

For product teams, this signals a shift.
▪️Behavioral testing is becoming as standard as QA.
▪️“Alignment health” will soon sit next to accuracy and latency metrics.
▪️Safety is moving from assumption to measurement.

And that’s Petri's real shift, turning “we think our models are safe” into something that can be tested and proven.

Petri vs Evals
Evals measure performance.
Petri measures behavior.

Evals answer, “Can it do the job?”
Petri answers, “Will it do it responsibly?”

That combination is what the next generation of AI product testing will look like.

Read the full breakdown here:

When AIs Audit Other AIs

If you're a PM trying to understand AI infrastructure and how to scale it, Madhumita Mantri's blog is for you.

Most AI PM content stays high-level, incluategy, use cases, and frameworks.

This one gets into the actual systems, decisions, and constraints.

In Part 1 of my interview with Madhumita Mantri, Staff PM at Walmart Marketplace, we discuss what it takes to build AI platforms that make decisions in under 200ms and still pass audits, roll back cleanly, and protect the business.

We talk through:
▪️Real-time trust & safety pipelines in production
▪️Infrastructure tradeoffs: Pinot, BigQuery, Flink, feature stores — and when to skip them
▪️How product priorities shift across industries like marketplaces, payments, analytics, and social
▪️What “data contracts” and SLOs look like when your model is always-on

This blueprint is worth studying for PMs building or preparing to build real, scalable AI systems.

Read Part 1: Real-Time PM Playbook: Trust, Speed & System Design

Real Talk AI (Part 1): Real-Time PM Playbook: Trust, Speed & System Design with Madhumita Mantri

Staff Product Manager@Walmart Marketplace | RealTime AI Decisions Platforms | Follow for Data, GenAI Products Podcast, Content | PM Interview Prep Coach | Ex-LinkedIn | StarTree | PayPal | Yahoo | Grace Hopper Speaker

Early in my career, I chased titles.
Not for power.
But because every new role felt like growth.

And to be fair, it worked.
Jumping from biology → QA → program management → product → AI product made me grow faster than if I had stayed still.
Each move gave me new perspectives and new skills.

But looking back, I also see the tradeoff.
When you move too quickly, you can gain range but miss depth.
You collect titles—but sometimes not mastery.

If I had to do it over again, I’d balance the two:
Patience to build depth.
Pivots to build range.

Because learning compounds.
And I’m still learning even now.

What helped you grow more in your career staying put, or making the leap?

Even after teaching AI Product Management for years, I still catch myself asking:
Should this be a Generative AI problem, or just classic NLP?

It’s not always obvious, which is why I wrote this week’s post.

Because while GenAI can do many of the things NLP does, that doesn’t always mean it should.

Generative AI gives you flexibility and fluency.

NLP gives you speed, structure, and predictability.

This post breaks down:
▪️What each approach is actually best at
▪️Where the trade-offs show up
▪️How to decide when you’re scoping an AI feature

Read the full post: https://lnkd.in/gvy39TRC

If you’ve ever overcomplicated a problem by reaching for GenAI too soon, you’re not alone.

#AIProductManagement #NLP #GenerativeAI #ProductThinking

Open article: When to Use Generative AI vs. NLP (And Why PMs Should Care) by aipmguru.substack.com

When to Use Generative AI vs. NLP (And Why PMs Should Care)

A job offer isn’t just about the role.
It’s about who you’ll be managed by.

Your hiring manager will shape your day-to-day experience, your confidence, and your long-term growth more than any job title ever will.

So don’t just prepare to answer questions in the interview.

Come ready to ask them.

Here are a few questions that reveal who they really are as a manager:

💬 "Can you share a time when someone on your team disagreed with you? How did you handle it?"
→ You’re testing for psychological safety, not just leadership style.

🧭 "What does success in this role look like after 6 months?"
→ Clarity = a good sign. Vague answers = 🚩.

🪴 "How do you support your team’s growth and career development?"
→ Great managers invest in people, not just projects.

🔄 "When was the last time you changed your mind based on feedback from your team?"
→ You’re looking for humility and adaptability.

🧠 "What’s your management style — and how do you adjust it for different team members?"
→ One-size-fits-all rarely works. Flexibility matters.

📉 "What’s something past team members have struggled with in this role or team?"
→ Transparency here = trust later.

🪞 "What’s something you’re working on improving as a manager?"
→ Leaders who are still learning tend to grow others, too.

Because here’s the truth:

You’re not just choosing a company.

You’re choosing a person who will have a direct impact on:

→ Your energy
→ Your growth trajectory
→ Your sense of purpose at work

So yes, evaluate the offer.

But interview your manager just as much as they’re interviewing you.

Ask the real questions.
Trust your gut on the answers.

What’s one question you wish you’d asked a past manager before accepting the role?👇

Know what kills AI adoption faster than bugs?

95% accuracy ✓
Amazing demos ✓ 
Strong initial adoption ✓

Then reality hits:

"Why was I rejected?"
"How did it pick this for me?"
"Can I trust this thing?"

Crickets from your AI.

Six months later, users bail for manual workarounds.

Here's the brutal truth: Accuracy without explanation is just fancy guesswork.

Your AI might be brilliant, but users will never fully trust it if they cannot understand its reasoning.

I just dropped a deep dive into fixing the transparency crisis in AI products. Spoiler: it's not about building better models.

Unlocking Trust with Explainable AI

AI can be a huge accelerator, but it can’t and shouldn't replace your brain.

I lean on no-code/low-code platforms and AI to move fast. But in the past, I burned almost 8 hours wrestling with a simple image issue that the “automation” just couldn’t handle.

The fix? Not another AI trick. Just common sense: check the code, delete the copies, rename the file, re-upload. Ten minutes later, the problem was solved.

AI is a great partner, but remember to apply your own logic. Sometimes, the most straightforward fix beats the most innovative tool and could save you a whole day.

Want your AI agents actually to solve real problems? Here’s the secret sauce: 

It’s all about understanding how people work, not just what they do, but why they do it. Build trust first. Show value step-by-step. 

Start small: define the core problem, create a simple MVP, then test with real users. Listen, adapt, and prove your AI’s worth. Once trust is established, scale up with more innovative features and deeper integrations.

The bottom line? Trust isn’t just a bonus; it’s the foundation for AI adoption. Without it, even the most innovative tech falls flat.

If you want your AI to make an impact, focus on the human story behind the data. Solving real problems isn’t just about algorithms; it's about people. 

Ready to build AI that truly sticks? Let’s make it happen. 

https://lnkd.in/gCDGzJyp

Jeffrey Shulman - thank you for the post and the chance to work with such thoughtful faculty.

What stood out most was how ready the UW Foster School of Business's marketing professors are to meaningfully incorporate AI into their teaching, not just as one assignment but as a skill students build over time.

The goal isn’t for students to use AI to finish assignments. It’s to help them critique it, think with it, and learn how to use it to make better marketing decisions.

This is exactly why I started Bluenox.ai: to help educators and organizations build lasting AI fluency together.

#AIinMarketing #HigherEd #TeachingInnovation #AIFluency #MarketingEducation #Bluenox

Professor at University of Washington - Michael G. Foster School of Business

How can professors prepare students for marketing careers that are changing rapidly as AI gets embedded more and more into the workflow?

Today, Shaili Guru , founder of Bluenox.ai taught the teachers. UW Foster School of Business faculty gathered for her excellent workshop. 

Shaili challenged us to rethink how we teach, not by adding one AI assignment, but by embedding AI fluency throughout the curriculum. 

I am proud to have created an opportunity for our faculty to learn how to ensure students don’t just use AI to finish assignments, but learn to how to harness AI to be better marketers. 

We came away from the session with teachable AI skills, ready-to-use prompts, and lesson plans.

Our department is eager and ready to prepare our students for the jobs of today and the world of tomorrow. Grateful that Shaili Guru contributed to our continuous improvement.  

To the other department chairs in my network, I encourage you to work with Bluenox.ai.

No alternative text description for this image

Anne Julliene (Joan) Barcelona, CISA, CISSP

Team culture isn’t always apparent in interviews.
But it’s never completely invisible.

In just 45 minutes, you can pick up on subtle signs if you’re looking for them.

Here’s how to spot whether a team is healthy, or just high-performing on paper:

✅ How do they talk about each other?
→ “We” vs. “they” tells you everything about trust and silos.

✅ Do they interrupt or build on each other?
→ A team that collaborates well in meetings often works well together outside of them, too.

✅ Is there laughter or just LinkedIn-level polish?
→ Authenticity usually signals safety.

✅ Do junior team members speak up or nod along?
→ Inclusion isn’t what leaders say; it’s who gets heard.

✅ When you ask about challenges, do they get real or go vague?
→ “We’re not perfect, but here’s how we work through it” = green flag.

✅ Do they seem energized — or just... polite?
→ People who like where they work show it.

Questions to ask during your interview:

1. “What’s something your team recently disagreed on — and how did you resolve it?”
2. “How do you support each other during crunch times?”
3. “What’s a recent win your team celebrated and how?”
4. “When team members have a tough day, what happens?”
5. “If I asked your team to describe the culture in 3 words, what would they say?”

Remember: the work you do matters.
But the people you do it with matter just as much.

Culture leaves clues. Don’t ignore them.

What’s one subtle signal you look for when evaluating team dynamics?👇

This is Part 2 of my interview with Madhumita Mantri, Staff PM at Walmart Marketplace.

In this post, we move from infrastructure to leadership and focus on AI PMs' skill sets to lead high-stakes, always-on systems across teams, metrics, and production workflows.

We cover:
▪️ How to structure DS/ML/DE/PM teams around decisions
▪️ Defining success with layered metrics: North Stars, guardrails, rollback plans
▪️ Real-time vs batch tradeoffs and how to explain them
▪️ What most teams underestimate when scaling from prototype to production
▪️ GenAI workflows, synthetic red teaming, and how she mentors AI PMs

This post is for PMs managing ambiguity, navigating cross-functional alignment, and leading AI responsibly at scale.

Read Part 2: Scaling Decisions, Leading Teams & GenAI Habits
https://lnkd.in/gVKVr4s2

If you missed Part 1, focused on real-time infrastructure, decision pipelines, and system-level tradeoffs, it’s here: https://lnkd.in/gq_Rb7tT

Real Talk AI (Part 2): Scaling Decisions, Leading Teams & GenAI Habits: Lessons from Madhumita Mantri

Remember when calculators weren't allowed in elementary school math class?

Teachers didn't ban them because they were "bad"; they held them back until we understood WHY 7 × 8 = 56, not just THAT it equals 56.

I keep thinking about this as I watch the AI-in-education debate unfold.

And honestly? I don't have all the answers. But I do have questions worth exploring.

The calculator comparison makes sense... until it doesn't.

Calculators did one thing: math. AI writes, reasons, creates, and collaborates. So maybe the foundations we need students to build are different now?

Here's where I'm torn:

One side of me thinks: Let kids struggle with writing essays and solving problems manually first. Those struggles build critical thinking muscles. If they lean on AI too early, do they skip the cognitive development that makes them effective thinkers?

The other side asks: Are we defining "foundations" based on a pre-AI world? What if prompt engineering, AI critique, and human-AI collaboration ARE foundational skills now? Are we preparing kids for a world that no longer exists?

And there's the timing problem:

Schools move slowly, even if we agree on a phased approach (foundations first, then AI). AI will have evolved three generations ahead by the time we roll out a framework. Are we always going to be playing catch-up?

What I think we can agree on:

Whether AI comes early or late, students need to learn to think critically, question outputs, and know when to trust their own judgment. The tool changes, but that skill doesn't.

The question isn't really "AI or no AI." It's "what does literacy look like in an AI-integrated world, and how do we build it?"

I don't have this figured out, but I think it's one of the most important conversations we can have right now.

What's your take? Should we protect the "foundation years" or rethink what foundation means? I'm genuinely curious where you land on this.

Even if the offer looks perfect on paper —
If your gut is telling you something’s off, listen.

Maybe the process felt rushed.
Maybe the manager dodged your questions.
Maybe you left the interview more drained than excited.

You can’t always explain it.

But your gut doesn’t need a spreadsheet.
It’s working with a lifetime of lived experience.

✨ Did you feel seen, or scanned?

✨ Did you feel energized or performative?

✨ Were your questions truly answered or just deflected with charm?

✨ Did their values appear in their behavior or on their website?

Here’s the truth:

You don’t have to accept an impressive offer that doesn’t feel right.

Not for the brand name.
Not for the paycheck.
Not to prove anything to anyone.

Because once you’re in...
You’ll be living that decision, day after day.

I’ve made this mistake before.

I talked myself into a role that looked amazing on paper…
But deep down, I knew it wasn’t the right fit.

And I wish I had listened to that voice sooner.

So this is your reminder:

If something doesn’t feel right, even when everything looks right
trust yourself.

You’re not just choosing a job.
You’re choosing who you’ll become.

Learn from my mistake and give your gut a seat at the table.

Have you ever walked away from an offer and felt relief instead of regret?👇

Taplio

Shaili Guru's Linkedin Analytics

Shaili Guru

Performance ScoreShaili Guru's LinkedIn Performance

Shaili Guru's Best Posts (last 30 days)

Ready to grow your LinkedIn presence?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Shaili Guru's Linkedin Analytics

Shaili Guru

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Shaili Guru's LinkedIn Performance

Shaili Guru's Best Posts .css-1emoo84{font-size:var(--chakra-fontSizes-md);font-weight:400;color:var(--chakra-colors-gray-600);}(last 30 days)

Ready to grow your LinkedIn presence?

Shaili Guru's Linkedin Analytics

Performance ScoreShaili Guru's LinkedIn Performance

Shaili Guru's Best Posts (last 30 days)