Resume Teardown #50: Undergrad with 3 Internships and AI Product Work Scoring 72% but Still Reading as Engineer
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 B.Tech undergrad at a top IIT with 3 internships (including one with a PPO offer building an AI call copilot and CRM platform), an electoral analytics project at real scale (2 lakh voters), and multiple hackathon entries scored 72%. This is one of the stronger student resumes we have seen. The gap is no longer "lack of experience." It is purely framing: the resume sells a builder-analyst when it could sell a product thinker.
The Resume
Background: B.Tech (Biotechnology) at a top IIT, graduating 2027. Three internships: forward deployment at an AI CRM startup (PPO offered), web developer at a space-tech startup, and data analyst at a customer loyalty company. Multiple technical projects including an electoral analytics platform, health sentiment analysis, AI stock intelligence, and blockchain voting. Active leadership: team lead for an international biology competition wiki, coordinator for Asia's largest student tech festival.
What looked good on the surface:
- Three real internships with shipped output, not just academic projects
- PPO offered at the most recent internship (employer confidence signal)
- AI product work: live transcript analysis, sentiment scoring, call copilot with sub-second latency
- Real-world scale: 0.3M calls analyzed, 2 lakh voters mapped, 52K health text records
- Hackathon performance: top 6 from 15,000+ registrations at a national competition
Score: 72%
What Makes This Resume Different from Most Student Resumes
Most student resumes we review have one of two problems: either they have only academic projects (no real users, no real stakes), or they have internships that were purely shadowing (no ownership, no delivery).
This resume has neither problem. It has:
- A real AI product shipped at a startup (CRM copilot with sub-second latency)
- A real analytics platform built for a political consultancy (2 lakh voters, 300 booths)
- A real contribution to a funding round (startup web platform delivered under investor deadline)
- Real scale numbers across multiple projects
The raw material for a 80%+ PM resume is here. The gap is entirely in how it is described.
The CRM Copilot: Product Work Described as Engineering
"Delivered agent guidance under 1 second by analyzing live customer transcripts, sentiments and objections"
This is a product decision disguised as an engineering metric. Sub-second latency is not a technical flex here. It is a product requirement: agents need real-time guidance during live calls, not after. If the model takes 3 seconds, the customer has already moved on and the guidance is useless.
But the bullet says: "I delivered latency." It should say: "I made a product decision about when guidance matters."
The PM reframe:
Before: "Delivered agent guidance under 1 second by analyzing live customer transcripts, sentiments and objections"
After: "Defined sub-second as the latency threshold for in-call agent guidance (agents lose the conversation window if guidance arrives late). Built a live transcript analysis pipeline that identifies customer sentiment and objections in real-time, enabling agents to course-correct mid-call."
Same work. But now the bullet shows: product decision (why sub-second matters), user context (agents lose the window), and capability delivered (course-correct mid-call). The engineering is present but serves the product story.
The Electoral Platform: Scale Without Product Logic
"Mapped 2 Lakh voters across 300 polling booths by building an interactive electoral map for a constituency"
This is impressive scale for a student project. 200,000 voters across 300 locations, built for a real political consultancy. But the bullet describes the output (a map) without the product thinking:
- Who used this map and what decision did it support?
- How did you decide what data to surface vs hide?
- Why an interactive map rather than a static report?
The PM reframe:
Before: "Mapped 2 Lakh voters across 300 polling booths by building an interactive electoral map for a constituency"
After: "Built a constituency analytics tool for a political consultancy to prioritize field deployment. Mapped 2 lakh voters across 300 polling booths, then used ML scoring to rank booths by outreach potential so field teams could focus limited resources on highest-impact areas."
Now the bullet shows: user (political consultancy), decision (prioritize field deployment), method (ML scoring for ranking), and outcome (focus limited resources on highest-impact areas).
The Funding Contribution: Big Claim, Weak Ownership Verb
"Contributed to $1M funding by translating startup metrics and product insights into pitch-ready web platform"
This is a bold claim. A $1M funding round is a real outcome. But "contributed to" is the weakest possible ownership verb. It could mean you wrote 3 lines of code on the investor deck website, or it could mean you single-handedly built the platform that convinced investors.
The problem with overclaiming: If you say "contributed to $1M funding" in a PM interview, the interviewer will ask: "What exactly was your contribution? Would the funding have happened without your work?" If the answer is "probably yes, I just built the website," the bullet damages your credibility.
The fix: Separate your direct output from the funding outcome. Own what you shipped. Let the reader connect the dots.
Before: "Contributed to $1M funding by translating startup metrics and product insights into pitch-ready web platform"
After: "Built the investor-facing platform under a fixed deadline, translating founder metrics and product traction data into an interactive web experience. Delivered in 10 weeks. The company closed a $1M round shortly after."
Now you own the work (built the platform) and the context is clear (investor deadline, metrics translation) without overclaiming causation.
The Resume's Structural Problem: Too Many Projects, No Hierarchy
This resume has 5 projects plus 2 hackathon entries plus 3 internships. For a recruiter spending 15 seconds, that is 10 things competing for attention with no clear signal of "read these first."
For PM applications, you need a visible hierarchy:
- Top 2 internships (CRM AI copilot, electoral analytics platform) should get the most space and the strongest PM framing
- Supporting projects (health sentiment, stock intelligence) can be 1-2 lines each showing breadth
- Remove or compress anything that does not show PM signal (blockchain voting, protein docking)
The hackathon entry (identity resolution, top 6 from 15,000+) is strong and should stay. But it needs user-problem framing: whose problem is cross-channel identity resolution? What business decision does journey stitching enable?
Dimension Scores
Skills & Tools: 73%
Strong technical execution supported by real bullets. Python, SQL, React, Django, ML, and AI tooling are all demonstrated through shipped work. The gap: PM craft terms (prioritization, requirements, discovery, experimentation) are absent from experience bullets. The skills are there in practice but invisible in language.
Experience & Background: 76%
Three internships with delivery evidence and a PPO offer is notably strong for an undergrad. The resume shows initiative beyond coursework and real employer confidence. The gap: no PM positioning anywhere. A recruiter cannot tell this person wants to be a PM vs a full-stack engineer vs a data analyst.
Leadership & Impact: 70%
Real ownership signals: shipped CRM features, led a 5-member team, delivered under investor deadlines. Scale numbers make the work feel substantial. The gap: all impact is framed as "I built X" or "I delivered Y" rather than "users experienced Z" or "the business moved by W."
Domain Expertise: 66%
CRM/sales, political analytics, health sentiment, biotech tools, and fintech. This is too scattered for a recruiter to know what you want to build in. Not a problem for the score (students get domain credit for breadth), but a positioning problem for applications.
ATS Readiness: 63%
Multiple issues: no Summary section, some internship dates missing, ATS-unfriendly formatting with orphaned fragments, and PM keywords weakly represented in experience bullets. The resume would parse reasonably for engineering roles but fail keyword matching for PM job postings because the vocabulary is engineering-first.
Key Takeaways
1. The PPO signal is your strongest credibility marker. When an employer offers you a full-time role after an internship, that is the ultimate validation of your execution quality. Make sure the resume makes clear WHY you earned it: what you shipped, what impact it had, and what decisions you made that a typical intern would not have.
2. Sub-second latency is a product decision, not just a technical metric. When you build real-time AI guidance for sales agents, the latency threshold is the product choice. Say why that threshold matters for the user workflow: if guidance arrives late, the conversation window is gone. That one sentence transforms the bullet from engineering to product.
3. Scale numbers need product context. "2 lakh voters mapped" is impressive but meaningless without: who used this, what decision it supported, and why your specific approach mattered. Scale is the proof. The product logic is the headline.
4. "Contributed to" is a credibility trap. If you cannot clearly own the outcome, own the work instead. "Built X under Y constraint, delivered in Z timeframe" is more credible than "contributed to [$big number]" because interviewers will always drill into the latter.
5. For PM applications, hierarchy beats breadth. Ten projects with equal weight is worse than 3 projects with deep PM framing plus 5 compressed one-liners. Give your reader permission to skim the bottom and focus on the top.
6. Position with a summary, not with a hobbies section. Replace the hobbies line with a 2-sentence PM-targeting summary. A recruiter learns nothing from "practice badminton regularly" but learns everything from "IIT undergrad targeting APM roles, shipped an AI call copilot with sub-second latency at [startup], built electoral analytics for 2L voters."
The Pattern
This resume represents the "strong executor, invisible PM" archetype. The candidate has more real product-adjacent work than 90% of student PM applicants: shipped AI features, real users, real scale, employer confidence (PPO). But the resume presents it all through an engineering lens. Every bullet leads with what was built. None leads with why it was built, who it was built for, or what changed after launch.
The fix is surgical, not structural. The experiences do not need to change. The framing does: add a PM summary, rewrite the top 3-4 bullets with user-problem-first structure, compress the tail, and drop the hobbies section. This resume is 2-3 hours of editing away from 80%+.
Score your own resume to see how your PM resume performs across all four dimensions.