Internships as Experiments: Compound your Career
Why I chose Apple > Deepmind & what internships teach you (that a resume never shows)
5 internships. 4 full-time jobs. 4 role pivots.
That’s my resume on paper. But what actually shaped it? A series of experiments - some worked, some I still second-guess but all designed to answer one question: Where can I solve problems that matter?
If I could give one piece of advice to anyone exploring a pivot or students:
"Don’t optimize gigs/internships for signaling. Use them to discover what actually matters to you."I chose internships on curiosity: skills I hadn't built, environments I hadn't seen.
The result? A path that looks non-linear but feels earned: Data Scientist → Growth & Recommendations PM (Ola, Amazon) → VC → Chief of Staff → AI PM (Apple, Samsung Research, Niantic Spatial).
Through-line? Turning frontier AI/ML into scalable products and businesses.
Each role was a pressure test, a hypothesis to validate. Some gave me conviction. Others ruled things out. But all helped me place sharper, intentional bets.
What that looks like: A few real experiments
Experiment 1: Finding the "Why"
Hypothesis: I want to use tech skills to make a positive impact grounded in data.
Test: Two very different internships - MuSigma (analytics consulting) and SocialCops (40-person data-for-good startup). At MuSigma, I worked on forecasting models for Fortune 50 enterprises. Clean data, polished presentations, impressed clients. At SocialCops, I built dashboards for the Indian government to identify school dropout rates. Messy data, scrappy solutions, enabling policy decisions.
What I Learned:
I cared more about creating direct real-world impact than polished outputs
I didn’t just want to code. I wanted to design systems, experiences and understand users.
The Shift: This clarity led me to Ola (Uber of India), where I grew from data scientist to PM— scaling rideshare from 0to1, improving maps ML and payments for 300M+ users, and helping economically uplift underserved drivers.
Experiment 2: Perspective from all sides of the table
Hypothesis: To be effective in business, I need to understand different sides of the table.
Test: Deliberately experimented at both ends of the strategic spectrum during MBA: venture capital (Stellaris) and Chief of Staff (Uniphore, a conversational AI startup).
In VC, I learnt how to evaluate businesses, ask the right questions to assess products/teams/companies, spotting founders early and most importantly, how rare real product-market fit is.
As Chief of Staff: Got a front-row seat to startup chaos & change management in wake of GenAI - M&A, OKRs, org design, AI platform pivots in customer service.
What I Learned: Being in both roles for 6+ months respectively allowed me to develop an org and market lens: from products to business models, from roadmap to runway. I started reading between funding rounds, strategic partnerships, and pitch decks, ruthlessly asking the fundamental question of: What actually makes a good business?
The Shift: This dual perspective now helps me build AI products that are not just technically strong, but commercially viable.
Graduating in 2023, I contemplated switching to either. But I knew: if there was ever a time to be an AI PM in Silicon Valley, it was now — a drivers seat to shape the right product, story and GTM loops for PMF of AI-native tech. As tweeted:
Experiment 3: The Hard Trade-off - DeepMind vs. Apple
Hypothesis: I thrive building AI experiences for real users.
Test: Summer 2022, I had an offer to intern as PM at DeepMind (pre-ChatGPT). I chose Apple's Visual Intelligence team instead. That decision still lingers in my head.
Trade-off:
DeepMind: Cutting-edge AI research, prestigious brand, future-focused
Apple: 0to1 opportunity to build AI for scale, production reality, user-focused
Would DeepMind have opened more doors in today’s GenAI-infused world? Possibly. But apple taught me what I needed most.
What I Learned: My edge is in the messy middle - translating AI research into product, and product into something people actually use. Ever used the sticker feature or copy-pasted text from an image on your iPhone? I helped envision that in 3D at Apple—building 0to1 AI experiences, designing data strategy, and learning what breaks in production.
Apple taught me the art of building truly AI-native products that feel like magic but also what breaks while scaling research to 1B+ users.
The Shift: That experience shaped how I now approach AI commercialization and gave me early reps in multimodal foundation models—before LLMs went mainstream.
What These Experiments Taught Me
What you just read was hypothesis-driven decision-making in action. I never had a "perfect" internship. But each gave me data points on:
what energizes me day-to-day
where I thrive
what problems I care about.
This clarity led me from data science → product → AI commercialization.
This is a story of disciplined exploration, of bets—some small, some scary, all curiosity-driven. I built range and a through-line by following my curiosity in AI while navigating functions and orgs.
Along the way, I learned:
Curiosity > Prestige: Sometimes the role you walk away from is what helps you define what you'll walk toward.
Friction beats comfort: Scrappy civic-tech or CoS chaos taught me more than polished brands.
Novel "what" > smooth "how": I do my best work on must-have use cases for frontier tech, in environments with agency and fast feedback loops
Informed conviction > perfect arc: A non-linear path is risky but may compound with intention.
Bridge building: I thrive translating between technical possibilities and market realities
At every stage, I continued to learn, listen, refine my conviction on where I create most value and what I want to bet on.
Takeaways for Anyone Pivoting & Students
Whether you're in undergrad, business school, or mid-career:
Treat internships like experiments. Test not just the work, but the environment, people, and problems
Pick them to stress-test your assumptions. Pick them for the friction, not just fit.
Leave with a clearer no, not just a glossier yes
Choose roles that teach you more than they signal.
Don’t over-romanticize someone else’s career path.
The internships that teach you how you work, not just how others work, are often the ones that change your arc completely. You don't need the cleanest arc, just informed conviction for your next bet.
Mine wasn't linear. It still isn't. But it's mine. And it's slowly making more sense.
Always experimenting, but with a lot more signal than noise.
Building something frontier and want to chat 0→1 AI PMF? Let's connect!
Solid advice. Follow your curiosity, always be learning.