Lessons from an 11-Month Job Search: How I Used AI to Land a New PM Role

A few weeks ago Block laid off 40% of its staff. At the same time OpenClaw has been taking Twitter by storm. Agentic Coding is all the rage. "Agents" has replaced "AI" as the buzzword of choice. Everyone who’s even adjacent to AI knows that the world is shifting and that a lot of what makes up traditional white collar work is rapidly being commoditized and replaced by AI. Looking for a new Product Mangement role right now is a bit of a high wire act. However, this might also be the last opportunity you’ll get to shift into something that’s more future proof and AI-compatible.

I’ve just moved to a new role myself. After 11 months of work spanning hundreds of hours of preparation and interviews with over 10 companies, I recently started a new role as a Technical Product Manager at Waymo working on the Simulation team. I’m excited to join Waymo, and the journey here taught me a lot about what preparing and interviewing is like in today’s job market.

I thought I’d share what worked, what didn't, and what I'd do differently. I’ve always been a huge skeptic of the PM Industrial Complex: the commoditization of Product Management interviewing into an SAT-Prep-like product (only $999 for the study book with all the answers!) even though the material itself is wildly detached from what makes a good Product Manager. I’m here to tell you that while you do need to prepare diligently for interviews, you probably don’t need to put yourself through a paid prep program — and AI is the reason why.

The Current Reality

Let me be direct: the job market for Product Managers is challenging right now. You'll need strong domain expertise and internal champions to find success. Of all my applications, I made it the furthest on roles where I had strong referrals from people who could advocate for me internally. A realistic timeline matters too — I started casually exploring in March 2025 and accepted my offer in February 2026. If you're planning a job search, expect 6+ months from start to offer even if you’re hustling.

When I felt ready to really double down on my dream opportunities, I focused on quality over quantity. This selectivity meant fewer applications but higher conversion rates.

My Competitive Advantage: An AI Interview Coach

As I reflect on this process, I’m fairly confident that the AI system I built up throughout this process was the differentiator that helped me stand out and stick the landing on key interviews. I built this system over time inside of Notion, ChatGPT, and Claude. By the end I had a refined workflow down that I’ll share with you here. To succeed in this job market you’ll need to be savvy about using AI to stand out. You’ll need to make an AI PM Interview Coach tailored specifically to you. Let me show you what I mean.

1. Start with Context

The first thing to build is your context corpus: the collection of data that you can let any AI model ingest and reference as they help you prepare. Without a robust, personalized corpus of data you’ll just get generic feedback and guidance.

I built my context corpus inside Notion. I started by brainstorming a list of stories, anecdotes, and experiences that I thought I might want to reference during my interviews. I then went through each item on the list and wrote out STAR bullets for each.

On top of this corpus of STAR stories, I added my resume, a bio generated by asking AI to summarize my LinkedIn, and personal preferences I wanted to document, such as my desire to work on technical, AI products. Finally, as I came across helpful advice online about PM interviewing, I stored that advice in this corpus as well to ensure that my AI Coach could critique and guide me while incorporating high-quality advice from outside its foundation model.

With this context corpus completed, I created a new Project inside Claude for the job hunt. I uploaded an export of all my content in Notion as source documents to this project. To make sure that Claude worked as a coach within this project, I added the following Project instructions:

I will use this project to prepare for interviews for PM job opportunities. Give me concise, extremely direct and honest feedback. Never be sycophantic. Never hallucinate or provide information that is generic. Always be objective and use knowledge directly relevant to the company or role at hand to answer the question specifically. Generic answers are not useful to me. If I am talking about a topic in regards to a specific company, draw on the context I have uploaded as well as context online and in your foundation model to give me advice that is hyper specific to that company. When you give advice, make clear why you are giving that advice for that company. Don't solve problems for me, write as a guide or coach to help me prepare myself.

2. Seed Unique Threads For Each Role

With a Claude project stood up, it was time to use it for actual applications. I created one conversation per role to allow context for each interview process to build over time, without polluting company-specific content across roles.

Next up I would use AI deep research to go deep on the company at hand. Whenever I was starting a new interview process, I would ask ChatGPT to write a Deep Research memo for me on that company to help me understand how the company operates internally, and to guide and seed future AI guidance. Here was the prompt I would use:

I need you to prepare a memo about <COMPANY X>. My goal is to understand deeply how the company speaks internally so that I can mirror that language during my interviews. Answer the following questions:

- What is the company's most pressing challenges and current focus?
- How is the company approaching those challenges? How are its strategies unique from competitors?
- What terminology and phrases are used most commonly by executives in public communications?
- What are the core product principles and philosophies of the company?

Refer to online sources such as blog posts, earnings reports, podcast transcripts, press releases, and anything else you can find to answer these questions. Prioritize heavily how representatives of the company itself speak, not how the company is discussed by third-parties such as reporters.

I exported these memos from ChatGPT and provided them as context to my interview thread in Claude. I would also provide the job description for the role.

3. Prepare a Study Plan

With the Claude thread seeded with Project-level and Company-level context, I was ready to create a study plan. I would ask Claude to create a study plan for me that matched the amount of time I had before my first interview. This would typically entail reading up on the company’s blog, reviewing the memo generated by ChatGPT, and doing some light mock case practice. Claude would prepare a day-by-day study plan for me that I would follow.

4. The Interview Prep → Feedback → Prep Loop

Finally, it was time to start interviewing. I would use the following process to prepare for and reflect on each interview:

Create Your Initial Cheat Sheet

For the first interview, usually a preliminary screen, I would ask Claude to prepare a cheat sheet for me. Over time, I landed on the following prompt and cheat sheet template which worked well:

Prepare a cheat sheet for me for my upcoming interview. The interview is with <PERSON; ROLE> and will focus on <FOCUS>. Consider all of the context available to you in this project, and in this chat. Your cheat sheet should be personalized to myself, this company, this role, and this interview. You should generate content for the cheat sheet by selecting content from Project context that best fits this company. Write only in bullet points in a glanceable format. I will skim and glance at this cheat sheet during my interview. Follow this outline for the cheat sheet:

- Suggested introduction
- Key strategic points to keep in mind (What is this interviewer looking for? How do I win this interview?)
- Key changes to keep in mind based on prior feedback
- Stories to keep top of mind
- Key numbers from selected stories
- Questions to ask

I would keep the generated cheat sheet handy during my interview to help me stay on topic.

Reflect On Your Performance

During my interview I would either ask the interviewer permission to use an AI note-taker, or I would take detailed notes about the flow of the interview, what was asked, and how I answered each question. Ideally I would end the interview with a full transcript of it, if permission to do so was granted.

Immediately following each interview, I would provide these notes and the transcript to the Claude chat for that role and ask:

Based on these notes from the prior interview, generate actionable feedback that I need to keep in mind for my future interviews. Focus on specific things that I need to change to improve my performance.

Repeat

With Claude’s feedback on my interview in hand (often harsh, thanks to my Project-level instructions), I would ask Claude to generate a cheat sheet for my next interview — ensuring it incorporated feedback from the prior interview into the next cheat sheet. Often I was able to complete this notes → feedback → cheat sheet cycle in the 10 minutes or so I had between interviews during onsites. For more spread-out interviews this schedule was obviously easier.

The result is that each cheat sheet was hyper-tailored to the up-coming interview and incorporated feedback from the prior interviews into it. Often for me that meant being reminded to not be verbose, and to start each response with a crisp answer before providing details.

Once I got this workflow well oiled I was able to get rapid, high-quality feedback on each interview — just like having a private coach. However the attention I had placed on providing personal context and helpful guidance meant that the feedback was actually unique, not just the generic feedback you read online when preparing for interviews.

This system mattered because it created consistency, personalization, and rapid improvement. Each interview made me better. But here's the key insight: AI was a force multiplier, not a replacement for real preparation. I still needed to read company materials myself, talk to real people, and genuinely understand the domain.

What Actually Worked

Alongside this Interview Coach that I built up, there were some additional aspects that I believe helped me.

Strong Referrals with Internal Champions

Not all referrals are equal. A casual LinkedIn connection submitting your resume certainly isn't enough. A close colleague on a different team likely isn’t even enough.

The final offers I got came through people who could actively advocate for me during the process and were on the team I was applying for. If you don't have someone willing to champion you internally, I wouldn’t expect the referral to move the needle meaningfully.

Domain Expertise

While not necessarily critical, having specific domain expertise definitely helped in several of my interviews. From what I can tell now more-so than ever hiring managers are trying to determine why someone is the best person specifically of many candidates for this specific role, not just a great candidate overall. To the degree that you can speak the language of the domain and team at hand — even if just from quick research ahead of time — it’s helpful. Selecting stories and anecdotes for behavioral questions that are tailored to a domain also often felt more powerful than “better” stories that weren’t as relevant to the domain.

Practice Interviews

My early applications were learning opportunities — and I bombed some of them. But each conversation made me better. Mock interviews have value, but actual company interviews are 10x more useful. Every company is different, every process is mercurial. Get comfortable with the randomness by doing it for real. I never accepted an interview invite for a company I wasn’t genuinely interested in. But if you start your preparation with interviews at companies that are your ideal targets you can get shots on goal that might result in a great offer, or just good interview practice at the very least.

Deep Company Research

Before every first interview, I read everything I could find: blog posts, product docs, conference talks, analyst reports. I researched interviewers on LinkedIn. I used AI to extract company language and values from public materials. Answers are often hiding in plain sight—you just need to look. There were more than a few case study questions I got that I was able to answer by drawing directly on content from the company’s own online publications.

Fast Feedback Loops

Being diligent about post-interview notes and submitting them to Claude with instructions to help me improve allowed me to identify patterns in my mistakes. I saw the traps I was falling into and course-corrected quickly. In this day of AI you can have something close to a personal coach at your beck and call. Just make sure you have prompted your Agent to be brutally honest. Improvement comes from brutal honesty, not self-congratulation.

Key Lessons and Advice

If I could restart my search knowing what I know now, here's what I'd want to know and what I’d change:

Start By Documenting Everything

Build your content corpus right away. Every notable project, decision, win, loss, or argument in your product career history. When you need to pull examples in interviews, you'll have them ready. This was one of the highest-leverage investments I made. AI allows you to parse through a historically unwieldy amount of material when prepping for interviews.

Invest in your AI system immediately. I wasted time doing generic prep before realizing that I needed a more systematized approach. I could have been building personalized materials from the start. Start there and then iterate and improve the system as you go through interview rounds. Your preparation system is your competitive advantage. Document your work history, build a content corpus, create company-specific materials. This is where AI shines—not in replacing preparation, but in personalizing and accelerating it.

Practice first with companies that aren’t your perfect match. Get comfortable by risking failure at companies that aren’t necessarily your dream. I would never recommend intentionally wasting the time of a company. If you aren’t serious about the role, don’t take an interview invite. Your lack of interest will derail you anyways. Instead, select companies that are interesting to you but aren’t your dream and start with applications there. Maybe you get lucky, fall in love with the company, and finish your process early. Or perhaps you brush off some cob webs, take an L, and emerge more ready for your next interview cycle.

Prioritize and prize opportunities with strong referrals and clear role fit. I wasted time on applications where neither existed. The conversion rate was near zero. If you have roles that are obvious fits for your background and have close internal referrals available to you, those might be good options to keep towards the back-end of your process when you’ve had some practice.

Be as harsh as you can on self-assessment. I prompted Claude to be mean in feedback sessions. "You did great" doesn't help you improve. Brutal honesty does. AI models are often sycophantic. Make sure they know that the best way to help you is by tearing you down.

Start preparing before you're desperate to leave. I started in May with a target start date of March—giving myself 10 months. This timing was right. If I'd started in August, I would have been crunched. Practice takes time, system-building takes time, and the market is slow. Start 6+ months before you need to leave.

Avoid the PM Industrial Complex. I went through this entire process with AI, free internet resources, and a Product Career Accelerator (PCA) free trial (which ultimately wasn't more helpful than some good blog posts). I didn't need expensive coaching or bootcamps. Strong work experience + right referrals + right opening = what matters. Save your money.

Final Thoughts

The market is tough, but systematic preparation works. The PM industrial complex will try to sell you expensive coaching and bootcamps. You don't need them. You need strong referrals, relevant experience, and a system for getting better with every interview.

I'm excited to start at Waymo working on simulation challenges — it's the most interesting problem space I could have found. But I also know I got lucky. I had strong advocates, relevant domain expertise, and the luxury of time to prepare properly.

If you're navigating your own search, I hope these lessons help. Build your system, get your reps, and be patient. The right opportunity will come.