The operating model
Growth
Content, strategy, and analytics for one startup at a time.
Growth is AINU's startup-facing operator team. The team embeds with one Boston consumer startup at a time and works the full loop: who the product is for, how it should sound, and which student-native channels actually move it.
ICP and positioning
Narrow in on the company's ideal customer profile and make the positioning and messaging unmistakably clear.
Student-native content tests
Short-form TikTok and Instagram experiments: hooks, campus-specific concepts, and creator-style distribution.
Growth engine and analytics
Build repeatable ways to put the product in front of the right users, measure downloads and engagement, and report learnings back to the founder.
Who it's for
Students interested in consumer startups, brand and content strategy, analytics, and practical AI workflows.
What you'll ship
- ICP and positioning brief
- Short-form content system
- Growth analytics report
- Portfolio-ready operator work
Investment
Institutional-quality memos on AI-driven mispricings.
Investment is a selective equity research team. Analysts look for public companies where AI has changed market expectations faster than the price — in either direction — and publish sourced, professional memos with a clear call, long or short.
AI repricing research
Find companies where AI has shifted market expectations — priced too bullish or too bearish — and build the evidence either way.
Investment memo writing
Sourced, professional memos with a thesis, valuation, catalysts, the opposing case, and explicit kill criteria.
AI as research infrastructure
Reusable research workflows: verified data, accelerated modeling, and judgment that stays with the analyst.
Who it's for
Students interested in public markets, equity research, valuation, diligence, and technical business writing.
What you'll ship
- Long or short investment memo
- Valuation and catalyst work
- Thesis with kill criteria
- Published research packet
Physical AI
Reinforcement learning, simulation, and sim-to-real demos.
Physical AI is the hands-on technical team. Students train policies in simulation, work on real manipulation problems, and connect modern models to hardware — building embodied systems you can watch run.
Robotic manipulation
LEAP Hand work, force control, gripping and cube reorientation, robot arms, and peg-in-hole or handoff tasks.
Simulation to reality
MuJoCo, domain randomization, behavioral cloning, and reinforcement learning — then sim-to-real transfer.
Vision-language-action systems
Connect modern AI models to physical hardware and hold them to measurable, real-world tasks.
Who it's for
Students who want technical depth in reinforcement learning, simulation, robotic manipulation, and vision-language-action systems.
What you'll ship
- Working robotics demo
- Sim-to-real experiment
- Open-source contribution
- Technical writeup or showcase artifact