Resume Teardown #49: IIT Dual-Degree with Elite AI Projects but Zero PM Language
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 dual-degree student (B.Tech + M.Tech) at a top IIT with a 9.2 CGPA, a major automotive tech internship, a competitive US research fellowship, and 7 AI/ML projects scored 61%. The technical caliber is elite. The problem: every single bullet answers "what model did I build?" and none answers "what user problem did I solve?" For PM hiring, this resume is invisible.
The Resume
Background: B.Tech (Biological Engineering) + Integrated M.Tech (Data Science) at a top IIT. 9.2 CGPA. ML/AI intern at a major automotive technology company. Research intern at a top US medical school (competitive national fellowship, 1 of 75 nationally). Current thesis work at a premier Indian medical institute. 7 technical projects spanning healthcare AI, agricultural advisory systems, conversational AI evaluation, and financial ML. Multiple national-level awards and competitive selections.
What looked good on the surface:
- Elite academic credentials with a dual technical + data science degree
- Real internship at a global automotive tech company (built an AI platform, >95% classification accuracy)
- Research at a top US medical school through a competitive national fellowship
- Strong technical range: deep learning, NLP, RAG systems, MLOps pipelines
- National-level recognition (top 6 in a hackathon with 15,000+ participants)
Score: 61%
Why 61% With This Pedigree?
This is the central lesson of this teardown. A resume with a top IIT, a US medical school, a global auto-tech company, a premier Indian hospital, and a 9.2 CGPA scoring 61% for PM seems wrong. It is not wrong. It reveals the fundamental difference between "impressive person" and "PM-ready resume."
PM hiring managers do not score credentials. They score evidence of product thinking: user problems identified, decisions made, trade-offs navigated, outcomes measured from a user perspective. This resume has none of those words anywhere. Not one bullet starts with a user problem. Not one bullet names a product decision. Not one bullet shows a trade-off between what to build and what to skip.
The resume is technically excellent. It would score 85%+ for ML engineering or research roles. But for PM, technical execution is the Skills dimension (weighted 60% for students), and even there, PM craft is expected alongside technical depth.
The Auto-Tech Internship Bullet: Your Best PM Signal, Wrongly Framed
"Developed an AI-driven incident response platform, achieving >95% classification accuracy in PoC validation"
This is actually product work. An incident response platform has users (operations teams), a workflow (incidents come in, get classified, get routed, get resolved), and product decisions (what confidence threshold triggers auto-routing vs human review? What happens when the model is wrong?).
But the bullet says none of that. It says: "I built a model. It is accurate." That is an ML engineer's bullet.
The PM reframe:
Before: "Developed an AI-driven incident response platform, achieving >95% classification accuracy in PoC validation"
After: "Designed an AI-powered incident triage system for operations teams to reduce manual classification overhead. Defined routing logic with a >95% confidence threshold for auto-assignment and a human-review fallback for ambiguous cases. Validated with the ops team through a PoC that processed [N] incidents."
Same work. But now the bullet shows: target user (ops teams), problem (manual classification overhead), product decision (confidence threshold + fallback), and validation method (PoC with real users).
The Agricultural Advisory System: Architecture Without a Farmer
"Designed an end-to-end AI agricultural advisory platform using React, FastAPI, Claude API, and RAG for evidence-based crop recommendations"
The technology stack leads: React, FastAPI, Claude API, RAG. A PM recruiter's eyes glaze over. They do not care about FastAPI. They care: who uses this? What problem does it solve? Why this approach over alternatives?
The PM reframe:
Before: "Designed an end-to-end AI agricultural advisory platform using React, FastAPI, Claude API, and RAG for evidence-based crop recommendations"
After: "Built a crop advisory tool for small-holder farmers to get personalized, evidence-based recommendations. Used RAG over agricultural research papers so advice stays current without retraining. Developed a yield-risk model (87% ROC-AUC) to flag high-risk crop combinations before planting season."
Now the bullet answers: who (small-holder farmers), what problem (need personalized crop advice), why RAG (stay current without retraining), and what outcome (87% accuracy on risk prediction). The technology is still visible but it supports the product story rather than being the story.
The Pattern: Seven Projects, Zero Product Decisions
Every project on this resume follows the same structure:
- Built/Developed/Implemented [technology]
- Using [stack/tools/methods]
- Achieving [model metric]
This is the ML research paper format: method, implementation, results. It works perfectly for ML conferences. It does not work for PM hiring because it never shows:
- Who has the problem? (No user mentioned in any project)
- What decision did you make? (No trade-off, no scope choice, no prioritization)
- What changed for the user? (Only model metrics, never user outcomes)
You do not need to change what you built. You need to add the layer of thinking that already happened but was never written down: Why this project? Who benefits? What did you choose NOT to build? What would you do differently?
The Missing Summary: Researchers Get Binned as Researchers
Without a summary, a recruiter scans this resume and sees: top IIT (engineer), US medical school (researcher), auto-tech company (ML intern), medical institute (medical AI). They categorize you as an ML researcher. Tab closed.
The summary is your one chance to override that categorization:
Example: "Dual-degree student (B.Tech + M.Tech Data Science) at a top IIT targeting AI PM or technical APM roles. Built an AI incident response platform at a major auto-tech company (>95% accuracy), designed a RAG-based agricultural advisory system with 87% yield-risk prediction, and currently developing clinical AI for cancer screening at a premier medical institute. Interested in PM roles where technical product decisions shape user outcomes."
Three sentences. Now a recruiter sees: PM intent, product-adjacent experience, and technical credibility. They read the rest through a PM lens.
Resume Density: Too Much for PM Screening
This resume has 7 projects, 3 internships/research roles, 5 positions of responsibility, publications, scholarships, and extracurriculars. For a PM application, this is too much information without enough focus.
The rule for student PM resumes: A recruiter spends 15 seconds. They will read your summary and your top 2 projects. Everything else is background noise. If your top 2 projects are the most PM-relevant ones with user-problem-first framing, you pass the screen. If your top 2 are the most technically complex ones (which is how this resume is ordered), you get binned as engineering.
Ruthless prioritization:
- Keep: The auto-tech internship (reframed as product), Agricultural Advisory (reframed around farmer), and one more with user signal
- Compress: LexicHash, heart failure biomarker, portfolio management into one-liners or drop them for PM applications
- Add: A dedicated "Product Thinking" section with one PM case study
Dimension Scores
Skills & Tools: 64%
Technical toolkit is elite and supported by real project bullets. But PM craft is entirely absent: no evidence of prioritization, requirements, user research, roadmap thinking, or launch planning. For AI PM evaluation, the prompt correctly notes no PM-style decisions about model behavior, thresholds, or fallback flows.
Domain Expertise: 66%
Clear healthtech exposure through the medical institute thesis and US medical school research. Agriculture AI adds breadth. For a student, this is enough to signal domain interest. The gap: neither is framed as product-domain expertise (understanding user workflows in healthcare) vs research domain.
Leadership & Impact: 58%
Strong builder ownership. The auto-tech PoC and the agricultural system show end-to-end technical builds. TA for 100+ students and team leadership at Inter IIT show communication range. The gap: no product decisions, no user outcomes, no prioritization evidence.
Experience & Background: 55%
Elite credentials and internship brands. But no PM-adjacent framing anywhere. A recruiter cannot tell this person wants to be a PM. The structural gap: zero PM positioning despite impressive underlying experience.
ATS Readiness: 80%
Missing Summary header hurts keyword density at the top of the resume. PM keywords are almost entirely absent from experience bullets (only "stakeholder" and "metrics" appear weakly). The resume would parse well for ML/data science roles but fail keyword matching for PM job postings.
Key Takeaways
1. Elite credentials do not substitute for PM language. Top IIT, a US medical school, and a global auto-tech company will get you noticed in ML hiring. In PM hiring, credentials open the door only if the resume demonstrates product thinking. Without PM language, these brands get you binned as "great engineer, wrong role."
2. Model metrics are not user outcomes. "87% ROC-AUC" means something to an ML team. To a PM hiring manager, it means nothing unless you explain: what user decision does this accuracy enable? What happens at 80%? At 95%? The threshold choice IS the product decision.
3. For AI PM roles, the product decision around the model matters more than the model itself. What confidence threshold did you set? What happens when the model is wrong? How did you design the human-in-the-loop flow? These are the questions AI PM interviewers ask. Your resume should answer them before the interview.
4. Student resumes for PM need ruthless focus. Seven projects dilute attention. Two deeply-framed PM projects with user problems, decisions, and outcomes are stronger than seven technically impressive builds without user context.
5. A summary with PM intent changes everything. Without it, your resume gets categorized by your degree and most recent experience type. For a dual-degree in Biological Engineering and Data Science, that default is "researcher/ML engineer." The summary overrides that in 3 seconds.
The Pattern
This resume represents the "technically elite student with zero PM positioning" archetype. The candidate has built genuinely impressive AI systems, interned at world-class institutions, and demonstrated initiative at every level. But the resume is written for ML/research hiring, not PM hiring. Every bullet prioritizes the technology over the user, the model over the decision, and the accuracy over the outcome.
The fix is not to build new things. It is to describe existing things through a PM lens: who has the problem, what decision you made, why this scope, and what changed for the user. Add a targeting summary, reframe the top 2-3 projects around product decisions, and compress the rest.
The path from 61% to 75%+: summary with PM intent, auto-tech internship + agricultural project reframed around users and decisions, one project replaced with a PM case study or shipped product.
Score your own resume to see how your PM resume performs across all four dimensions.