Resume Teardown #26: Junior AI PM Who Shipped a GenAI Agent but Still Reads Like a Builder

This is part of our Resume Teardown series where we score real PM resumes (anonymized) and break down what the evaluation found.

TL;DR: A junior PM with an engineering background in JavaScript, React, and GenAI APIs scored 64%. The resume has one genuinely strong bullet (a GenAI support agent that automated 70% of inbound queries) and a fast promotion signal. But the overall read is "strong builder who does some PM work" rather than "PM who leverages technical skills." Experiment bullets lack outcomes, AI work lacks product judgment, and the consulting-era projects are framed as engineering accomplishments rather than product decisions.

The Resume

Background: Associate Product Manager at a D2C consumer brand (promoted from Product Trainee within 6 months, Aug 2025 - Present). Previously Product Intern at a product consultancy (Feb 2025 - Jul 2025) working on healthcare and sports betting apps. Before that, Product Management Consultant and Software Developer at an early-stage startup (Nov 2022 - Feb 2025). B.Tech in Mechatronics Engineering. AWS Cloud Practitioner certified.

What looked good on the surface: Shipped a GenAI support agent (70% query automation). Ran checkout funnel experiments. Got promoted fast. Technical fluency across React, LLM APIs, and analytics tools. Multiple shipped products.

Score: 64%

The Core Strength: You Actually Shipped AI

The Goodsy bullet is the best thing on this resume:

"Built and launched an in-house GenAI support agent that automated ~70% of inbound queries across email and chat, cutting average CX response time."

This is specific, quantified, and shows real ownership. For a junior PM, shipping an AI product that measurably changed a business metric is exactly the kind of bullet that gets a hiring manager's attention. It names the product, the technology, the scope, and the result.

The API cost reduction bullet from the healthcare internship is similarly strong: identifying an over-calling pattern and working with engineering to cut costs by 25%. It shows analytical thinking, cross-functional execution, and a measurable business outcome.

These two bullets alone justify a phone screen. The problem is that the rest of the resume doesn't maintain this standard.

The Core Problem: Builder Resume Wearing a PM Hat

The consulting-era section is where the resume loses its PM identity. Three of four bullets are pure engineering accomplishments:

"Built and shipped a live generative AI web app integrating Replicate and fal-ai for LLM and image generation"

"Launched a real-time driver drowsiness detection web app using Google MediaPipe 3D facemesh"

"Built and launched a production React.js site with Vite and automated CI/CD via Netlify"

These show you can build. They do not show PM judgment. A hiring manager reading these will ask: Who was the user? What problem were you solving? Did anyone use it? What did you learn? How did you decide what to build?

For AI PM roles specifically, hiring managers want to see product decisions about AI behavior, not just that you integrated an API. Did you evaluate output quality? Handle bad generations? Choose between model variants? Set confidence thresholds? The resume shows you can call LLM APIs. It doesn't yet show you can make product decisions about what those APIs should do.

The Experiment Gap

The D2C brand bullets mention "growth experiments" and "weekly experimentation cycles" but never land the outcome:

"Ran growth experiments across the checkout funnel (PDP, cart, OTP, upsell), working with design and engineering to lift conversion on hero SKUs."

What converted? By how much? Which experiment won and why?

"Led weekly experimentation cycles using PostHog and analytics to prioritise bets by projected revenue lift versus engineering effort."

What was the highest-impact decision this process led to? What changed because of your prioritization?

The experimentation infrastructure is clearly there. The results are missing. "Ran experiments" is activity. "Tested 3 upsell placements on PDP, found post-add-to-cart timing lifted AOV by 12%" is impact.

Company Context is Thin

None of the employers on this resume are household names. A recruiter scanning cannot immediately understand what these companies do, their stage, or the scale of the problems being solved.

The fix: One line per company.

• D2C brand: "D2C probiotic brand selling via Shopify, ~X monthly orders"
• Product consultancy: "Product consultancy building apps for healthcare and sports betting clients"
• Early-stage startup: "Early-stage startup building AI-powered interview prep tools"

That single line per employer resolves ambiguity and gives the hiring manager a mental model for the scope of your work.

Dimension Scores

Leadership & Impact: 68% The GenAI agent (70% automation) and API cost reduction (25% savings) carry this score. Several other bullets show ownership but stop at operational improvements without connecting to a business metric. For a junior PM, this is solid but could be 75%+ with outcomes added to the experiment bullets.

Skills & Tools: 67% Strong technical fluency for a junior PM. Experimentation, analytics (PostHog), PRD/BRD documentation, and cross-functional execution are all demonstrated. But the AI skills are only partially shown in PM terms. The resume names LLM APIs and prompt design but doesn't show eval frameworks, quality metrics, fallback design, or collaboration with ML specialists.

Experience & Background: 61% Clear progression from technical building into formal PM roles, with the fast promotion strengthening the narrative. But the role sequence is compressed (multiple short stints close together), and company context is thin. A hiring manager may question depth of ownership without tighter framing.

Domain Expertise: 58% Credible D2C e-commerce exposure through the current role (PDP, cart, upsell, subscriptions, retention). But the AI domain depth is limited to GenAI API integration without evidence of domain-specific AI challenges like hallucination handling, evaluation quality, or model selection decisions.

ATS Readiness: 89%

Strong ATS performance. Standard headers, consistent dates, ATS-friendly formatting. Minor warning on unexpanded acronyms (PDP, OTP, BRD, PRD, GTM, CX). PM keywords are well-represented: roadmap, stakeholder, metrics, prioritised, cross-functional, user research, retention, experimentation, requirements, go-to-market.

The 4 Changes That Would Move This Score

1. Add outcomes to every experiment bullet.

Before: "Ran growth experiments across the checkout funnel (PDP, cart, OTP, upsell), working with design and engineering to lift conversion on hero SKUs."

After: "Ran 12 growth experiments across the checkout funnel over 3 months. Top winner: moving upsell placement from PDP to post-add-to-cart lifted AOV by [X%] on hero SKUs."

Pick your top 2-3 experiments and add the result. Conversion lift, revenue impact, or validated learning.

2. Reframe the consulting-era projects as PM work.

Before: "Built and shipped a live generative AI web app integrating Replicate and fal-ai for LLM and image generation"

After: "Defined requirements and shipped an AI image generation tool for [user type]. Evaluated output quality across 3 model providers, chose Replicate for [reason]. [X] users in first month."

Show the product decision, not just the implementation. Who was it for? How did you evaluate whether the AI output was good enough? What did users do with it?

3. Add company context lines.

One sentence per employer explaining what they do, who they serve, and their approximate scale. This is a 5-minute fix that meaningfully changes how a recruiter reads the rest of your bullets.

4. Show AI product judgment on the GenAI agent.

Your strongest bullet (70% automation) can be even stronger. Add one sentence about the product decisions: How did you handle bad responses? What was the containment rate? Did you set quality thresholds? How did you decide which queries to automate vs. route to humans?

That's the difference between "I integrated an LLM" and "I made product decisions about an AI system." For AI PM roles, this distinction matters more than anything else on the resume.

The Pattern

This resume represents a common junior PM archetype: the technical builder who moved into product. The engineering skills are genuine and valuable. The shipped products are real. The promotion signal is strong.

The gap is framing. Every bullet needs to answer "what product decision did I make?" rather than "what did I build?" The Goodsy bullet proves this person knows how to write a strong PM bullet. The rest of the resume just needs the same treatment.

The path from 64% to 75%+:

• Add experiment outcomes (the data exists, it just isn't on the resume)
• Reframe builder bullets as product decisions
• Add company context for non-obvious employers
• Show AI product judgment, not just AI implementation

The raw material is strong for a junior PM. The resume just needs to consistently tell a PM story instead of alternating between PM and engineer.

Score your own resume to see how your PM resume performs across all four dimensions.