TNT Fund I
This deck is confidential. Enter the access code to continue.
This deck is confidential. Enter the access code to continue.
We built the community for MIT and Harvard
startup founders. Now we're investing in it.
Confidential - For Qualified Investors Only
Overview
The Market
The best venture returns come from the best founders. MIT and Harvard produce a disproportionate share.
| Metric | MIT/Harvard | Source |
|---|---|---|
| Unicorn likelihood | 50-60% higher than avg. VC-backed founder | Strebulaev, Stanford GSB (2025) |
| Share of VC-backed unicorns | #2 and #3 globally (behind Stanford) | PitchBook / Crunchbase |
| Pre-seed / seed capital share | ~25% of all early-stage capital | Beta Boom Pedigree Study |
Investment Thesis
120 pre-seed investments into the best AI and deep tech founders at MIT and Harvard.
$150K for 3% equity + $150K SAFE. Companies with moats that can't be commoditized by AI.
MIT and Harvard. #2 and #3 globally for unicorn founders.
AI infrastructure. Vertical AI for regulated industries. Deep tech with real IP.
$150K for 3% + $150K SAFE. 120 companies. First check in.
Embedded inside labs and classrooms where companies are born.
Operators, not suits. Founders work with us because we actually help.
The best founders come to us because they've heard what TNT does.
Proprietary AI tools to screen, score, match mentors, and track portfolio.
How It Works
We're building TNT into the brand that the best MIT and Harvard founders want on their cap table. That reputation is both our sourcing engine and our selection advantage.
Track Record
Breakout companies from our first three cohorts, all before we had a fund.
AI-powered full-stack app builder for entrepreneurs and agencies
Agent infrastructure with novel model system co-design
Quantum cameras for earth and space intelligence
AI-native engineering services holding company
Building the most power-dense energy system
Automates deal execution for investment banks
AI back-office for industrials and manufacturing
Premium private agent automation platform
Representative selection. Does not include all cohort companies.
Social Proof
We don't just invest. We show up. Founders go to bat for us because we've gone to bat for them.
Team
Founder & General Partner
Founded TNT at MIT and built it into the largest founder community across MIT and Harvard, with 180+ applications per cohort, 30+ companies supported across 3 cohorts, $20M+ raised by portfolio. Embedded on campus since day one: not networking from the outside, but building from inside the labs, classrooms, and dorm rooms where companies start. Deep relationships with faculty, student orgs, and the broader Boston startup ecosystem.
Co-founder & Fmr CEO, HubSpot ($35B)
CEO, Ginkgo Bioworks (MIT)
Current MIT and Harvard students who serve on the TNT team. They're our eyes and ears on campus, connected to every lab, club, and dorm room where the next great company might be forming.
The Program
10-week accelerator culminating in Demo Day. $300K ($150K for 3% equity + $150K SAFE).
Anthropic, Google Cloud, AWS, NVIDIA, and 25+ partners.
Incorporation, SAFEs, and IP protection
Experienced founders and operators matched to each team
Past and upcoming: CEOs of Formlabs, Liquid AI, Boston Dynamics, Delve, Kayak, and more
Direct introductions and fundraising support
In front of a crowd of 300+ during Boston Tech Week
Corporate Pilot Brokering: We're exploring partnerships with enterprises who pilot technology from our cohort companies, giving founders their first customers and partners real-time access to cutting-edge AI.
Timing
30+ companies, $20M+ raised, 3 cohorts, all bootstrapped. A fund formalizes what's already working.
180+ applications this semester alone. We're turning away good companies because we can't support them all.
Pre-seed valuations at MIT/Harvard are still reasonable ($8-20M). As AI hype inflates later-stage prices, early entry becomes more valuable.
With fund capital, we can secure significantly more credits and resources from partners like Anthropic, Google Cloud, AWS, and NVIDIA, plus dedicated mentors, legal, and Demo Day.
Top accelerators and VCs are converging on MIT/Harvard AI deal flow. The market is telling you where the best founders are. We're already there.
The Fund
$150K buys 3% at ~$5M implied. $150K uncapped MFN SAFE converts at next priced round (~$20M). Blended ~3.75% ownership at ~$8M effective, well below typical MIT/Harvard pre-seed valuations because founders choose us for the community, not just the check.
Why this matters: YC takes 7% for $500K. We take 3% for $300K, giving founders more capital per point of dilution than any comparable program. Our terms are designed to complement YC, not compete: founders can do both. Lower dilution means better founders say yes, which means better deal flow for LPs.
Every dollar goes to new companies at entry price. No capital diverted to markups. This maximizes the number of shots on goal: 120 companies instead of fewer bets with follow-on reserves. Eliminates signaling risk.
TNT Select Fund is a separate, future vehicle (like YC's Continuity Fund) that will follow on into Fund I winners at later stages. Fund I LPs get first access.
The Terms
We're an accelerator that helps startups with capital, mentorship, legal, and Demo Day. Our terms are the most founder-friendly in the market. And if a founder also does YC, the combined 10% dilution is still less than most single pre-seed rounds.
Deployment
| Year 1 | Year 2 | Year 3 | Total | |
|---|---|---|---|---|
| Cohorts³ | 15 + 15 | 20 + 20 | 25 + 25 | 6 cohorts |
| Companies | 30 | 40 | 50 | 120 |
| Invested | $9M | $12M | $15M | $36M |
2% on committed capital during investment period (yrs 1-3), 2% on invested capital post-investment (yrs 4-10). Fees cover salary, operations, Demo Day, and team.
³ Cohort sizes are targets based on current capacity. Actual sizes may vary depending on the quality and volume of applications in each cycle. Our current cohort is 20 companies. We invest in quality over quantity and will only scale cohort size as the applicant pool supports it.
Every dollar at entry price. No capital diverted to markups. Eliminates signaling risk.
Dedicated follow-on vehicle for Fund I winners. SPV co-invest for LPs.
Fund Economics
120 companies × $300K. 3% equity + uncapped SAFE. Returns adjusted for dilution across funding rounds.
| Outcome | Exit | Stake | Conservative | Base | Optimistic | Outlier |
|---|---|---|---|---|---|---|
| Fail | $0 | — | 55% → $0 | 46% → $0 | 35% → $0 | 35% → $0 |
| Survive | $10M | ~3.0% | 25% → $9M | 25% → $9M | 26% → $10M | 26% → $10M |
| Solid Exit | $50M | ~2.4% | 12% → $17M | 16% → $24M | 20% → $29M | 20% → $29M |
| Strong Exit | $150M | ~2.0% | 6% → $23M | 8% → $29M | 11% → $39M | 11% → $39M |
| Breakout | $500M | ~1.8% | 1% → $10M | 3% → $30M | 5% → $54M | 5% → $54M |
| Unicorn | $1B | ~1.6% | 1% → $17M | 2% → $35M | 3% → $52M | 2% → $48M |
| Decacorn | $10B | ~1.0% | n/a | n/a | n/a | 1% → $100M |
Ownership diluted per round: seed ~20%, A ~20%, B ~15%, C ~12%. Outlier assumes additional D/E rounds (~1.0% remaining). SAFE converts at seed (~$20M pre-money).
¹ Conservative uses industry-average failure and exit rates. MIT alumni companies survive at 80% over 5 years vs. 50% nationally (Roberts, Murray & Kim, MIT Sloan, 2015). MIT/Harvard founders also raise follow-on capital at significantly higher rates (Crunchbase, 2023). Actual outcomes for MIT/Harvard founders are likely better than conservative.
² Median seed-stage fund returns ~1.5-2.0x (Carta VC Fund Performance, 2025). Our conservative case of 1.7x reflects average industry outcomes before MIT/Harvard selection advantage.
The Platform
Fund I is a discovery engine. The TNT Select Fund is a separate follow-on vehicle (like YC's Continuity Fund) that doubles down on winners using insider data. Separate because follow-on has a different risk/return profile. Fund I LPs get priority access.
| Stage | Investment | Fund I Only | Fund I + Select |
|---|---|---|---|
| Pre-seed (Fund I) | $300K | 3.75% | 3.75% |
| After Seed dilution (20%) | — | 3.00% | 3.00% |
| Seed follow-on (Select) | $500K @ $25M | — | 5.00% |
| After Series A dilution (20%) | — | 2.40% | 4.00% |
| Series A follow-on (Select) | $1M @ $62.5M | — | 5.60% |
| After B/C/IPO dilution | — | 1.62% | 3.77% |
| Return on $1B exit | $1.8M total | $16M | $38M |
The Ask
TNT Fund I
Minimum LP commitment: $250K • Target close: Q3 2026
Confidential. For qualified investors only. Past performance does not guarantee future results.
Appendix 1
Fund I is built for maximum shots on goal at entry price. Reserving capital for follow-on dilutes the core strategy. Winners get followed on through a separate continuity vehicle (Fund II), giving LPs the option without dragging Fund I returns.
The cohort model does the heavy lifting. Companies support each other. Our programming (mentors, Demo Day, build nights) scales across the cohort, not 1:1. We're building a community, not a consulting firm. The best accelerators (YC, Techstars) prove this model works at scale.
Concentration is the strategy. These two schools produce more unicorn founders than any others. Owning this niche completely is more valuable than spreading thin across 20 schools. Expansion is a Fund II+ conversation once the brand is proven.
The brand lives on campus through the community we've built, which grows with each new cohort. The fund operates independently. Being alumni of MIT/Harvard actually deepens the network over time. Brian Halligan and Jason Kelly aren't students either.
Emerging managers consistently outperform established funds in pre-seed and seed. Smaller fund size means discipline. Proprietary access means better entry prices. And Fund I economics are the best LPs will ever get from this team.
Most VCs source from the outside through cold outreach and referrals. We source from the inside through events (fireside chats, hackathons, build nights, co-founder matching) that we've been running for years. By the time a team applies, we've already seen them build. That's intel no outside investor can replicate.
Appendix 2
| Use | Amount | % of Fund |
|---|---|---|
| Company investments (120 x $300K) | $36,000,000 | 80.0% |
| Management fees (10yr total) | ~$6,100,000 | 13.6% |
| Fund expenses (legal, audit, admin) | ~$1,000,000 | 2.2% |
| Reserves | ~$1,900,000 | 4.2% |
| Total | $45,000,000 | 100% |
Appendix 3
Ownership % drops through funding rounds, but stake value only goes up. Smaller slice of a much bigger pie.
| Stage | Round | Pre-Money | Dilution | Ownership | Stake Value |
|---|---|---|---|---|---|
| TNT entry (equity) | $300K | $5M | 3.0% | $150K | |
| SAFE converts at seed | $20M | 3.75% | $750K | ||
| Seed closes | $5M | $20M | ~20% | ~3.0% | ~$750K |
| Series A | $12M | $50M | ~20% | ~2.4% | ~$1.5M |
| Series B | $30M | $150M | ~15% | ~2.0% | ~$3.5M |
| Series C | $75M | $400M | ~12% | ~1.8% | ~$8.2M |
| IPO / Late | Varies | $1B+ | ~10% | ~1.6% | $16M+ |
Ownership %
Stake Value
Intimate conversations with world-class founders and operators. Past and upcoming speakers include CEOs of Formlabs, Liquid AI, Boston Dynamics, Delve, and more. These aren't panels. They're real conversations in small rooms.
We host hands-on building events where we see founders in action. This is where we spot technical talent and watch teams form in real time.
Current MIT and Harvard students serve on the TNT team. They're our eyes and ears on campus, connected to every lab, club, and dorm room where the next great company might be forming. This is structural, not networking.
We help technical founders find business co-founders and vice versa.
Applications are screened using TNT OS, our proprietary AI-powered evaluation platform. We score across four dimensions: Team, Thesis Fit, Idea Quality, and Execution Readiness. Final decisions are made by the GP after deep-dive interviews.
AI infrastructure, vertical AI for regulated industries, deep tech with real IP. We run every application through a defensibility filter: can a frontier model replicate this in 12 months? If yes, we pass. We want companies with proprietary data, domain expertise, or technical moats.
Teams get speakers, investor prep, investor intros, dedicated mentors, and build nights. Plus $300K+ in partner credits (Anthropic, Google Cloud, AWS, NVIDIA). The program builds toward Demo Day in front of 300+ investors during Boston Tech Week.
We don't disappear. TNT alumni stay in the network. We continue making VC intros, helping with follow-on raises, and connecting teams with each other. The cohort is a lifetime network.
MIT and Harvard rank #2 and #3 globally for producing unicorn founders (behind only Stanford). Together they've produced the founders of Moderna, HubSpot, Cloudflare, Dropbox, Akamai, and dozens more billion-dollar companies. The density of technical talent, research labs, and entrepreneurial culture is unmatched on the East Coast.
Both campuses sit within a mile of each other in Cambridge, MA. This proximity creates a unique cross-pollination of ideas, co-founder matching, and shared resources that no other university pairing can replicate.
Others have tried to build around this ecosystem, but no one has combined community, accelerator, and fund into a single integrated model the way TNT has. YC and other accelerators draw from a global pool. TNT is purpose-built for this niche.
The picks-and-shovels layer. Companies building the tooling, data pipelines, and agent frameworks that every AI application depends on. These moats can come from years of deep research or from scrappy teams who move fast enough to own a critical layer before anyone else. Either way, the result is network effects and switching costs that grow over time.
Healthcare, defense, finance, government, manufacturing. Industries where domain expertise, compliance requirements, and data access create natural barriers to entry. A general-purpose AI can't walk into a hospital or onto a factory floor.
Quantum sensing, nuclear energy, novel materials. Companies where years of PhD research create defensible intellectual property. These are hard to start, hard to replicate, and exactly what MIT and Harvard labs produce.
$150K for 3% equity plus $150K uncapped SAFE. Total $300K deployed per company. Effective entry at ~$8M (assuming a $20M seed), well below market pre-seed. The SAFE converts at the next priced round, giving the fund additional ownership as the company grows.
3% is less than half what YC takes (7%). The SAFE provides runway without additional dilution at entry. Founders who also do YC end up with ~10% combined dilution, still less than most single pre-seed rounds.
Entry at ~$8M effective valuation (assuming $20M seed) vs. $15-25M market pre-seed. The SAFE participates in valuation expansion. 120 companies means maximum shots on goal. First check in means the best possible entry price.
TNT Fellows are current MIT and Harvard students embedded across campus. They're in the labs, the dorms, the study groups where the next great company might be forming. This gives us structural visibility into founding teams months before they start fundraising.
Most VCs meet founders at the pitch stage. We meet them at the idea stage. Our fireside chats, hackathons, and build nights let us watch founders work before they ever apply. By the time they're ready to raise, we already know who they are.
We run events, connect founders, make intros, and provide resources long before we ever invest. This builds a reputation that no cold outreach or brand marketing can replicate. When founders are ready to raise, TNT is already on their shortlist.
Every founder who goes through TNT becomes part of a permanent network. Alumni refer new applicants, mentor current cohorts, and collaborate on projects. This creates a self-reinforcing cycle of trust that compounds over time.
180+ applications for our Spring 2026 cohort with zero paid marketing. Every applicant found us through the community, a friend in a previous cohort, or an event they attended. This is the highest-quality deal flow you can get: warm, pre-filtered, and self-selecting.
Each cohort strengthens the brand. Alumni tell their friends. Fireside chat attendees become applicants. The more founders we help, the more founders want in. This is a compounding advantage that gets stronger every semester.
AI-powered evaluation that scores every application across multiple dimensions: team strength, thesis fit, idea quality, and execution readiness. This lets us process 180+ applications with consistency and speed while surfacing the strongest candidates for GP review.
Automated matching between cohort companies and our mentor network based on industry, stage, and specific needs. The right mentor at the right time can change a company's trajectory.
Real-time dashboards for fundraising progress, KPIs, and milestones across the entire portfolio. Gives LPs transparency and helps the team identify which companies need attention.
Solid generates production-ready, full-stack web applications using React, Node.js, TypeScript, and Postgres, outputting clean, readable code rather than throwaway prototypes. Unlike many AI code tools, Solid produces maintainable codebases that developers can extend, deploy anywhere, and hand off to engineering teams. Well-suited for agencies prototyping for clients and founders shipping MVPs fast.
Subconscious is building infrastructure for deploying and scaling AI agents, distinguished by its co-design of model and runtime. Rather than layering orchestration on top of third-party LLM APIs, the team built a custom inference engine that handles task decomposition, tool use, and long-horizon reasoning natively. The technology emerged from MIT research and addresses core agent failure modes like brittle multi-step workflows and context window limitations.
Diffraqtion develops quantum camera systems that use photon-counting sensors and proprietary AI algorithms to extract significantly more information from incoming light than standard sensors, enabling dramatically higher resolution and faster processing than conventional optical surveillance. DARPA awarded the company a Direct-to-Phase II SBIR contract, and they are working with the Space Force's Space Domain Awareness Lab. Plans to launch their first satellite (Galileo-1) in 2028.
Cortex acquires and transforms engineering R&D services firms by deploying AI-powered knowledge capture and workflow automation across their operations, targeting sectors like autonomous vehicles, aerospace, and semiconductors where engineers repeatedly recreate complex work manually. Founded at Harvard Business School, the thesis is that configurations and processes that historically took weeks can be reduced to minutes, making each project accelerate the next.
Apollo Atomics is developing compact nuclear reactor technology backed by 15 years of research at MIT, designed to be significantly smaller than traditional pressurized water reactors while targeting electricity costs as low as 3 cents/kWh through higher power density, passive safety, and reduced plant complexity. Founded in 2024 and headquartered in Cambridge, MA, the company is backed by Y Combinator.
Maywood automates the most labor-intensive parts of M&A deal execution, automatically generating CIMs, presentations, financial models, and diligence responses, with data updates propagating throughout the entire system from first document upload to close. Founded by former Blackstone and BCG professionals with MIT engineering backgrounds, the company targets middle-market and large investment banks running multiple sell-side processes per year.
Zaqa automates repetitive back-office tasks for medium-sized manufacturing and distribution businesses, processing purchase orders in seconds versus minutes in legacy ERP systems like SAP. Founded out of Harvard Business School, the company emphasizes seamless integration with existing software to avoid painful multi-year migrations. Backed by Flybridge, IM Ventures, and Meridian.
ZeroClaw Labs is building a premium, privacy-first AI agent platform that lets enterprises deploy autonomous agents entirely on their own infrastructure. Founded by a Harvard team, ZeroClaw sits at the intersection of two massive trends: enterprise AI adoption and data sovereignty requirements. As every company races to deploy AI agents, the ones in regulated industries and security-conscious environments need a solution where nothing leaves their walls. ZeroClaw is that solution.
Rob founded TNT at MIT Sloan and built it into the largest founder community across MIT and Harvard, entirely bootstrapped. Over 3 cohorts, TNT has supported 30+ teams that have collectively raised $20M+, with sponsors including JPMorgan, Anthropic, Google, NVIDIA, and AWS.
Before MIT, Rob was Director of Strategy & Innovation at Thinktiv, where he led AI strategy, product strategy, and GTM design for B2B SaaS and PE-backed companies. Prior to that, he was a consultant at Deloitte, specializing in financial modeling, systems design, and research for federal and financial services clients.
Rob holds an MBA from MIT Sloan (Class of 2026), an MA in International Business, and a BS in Finance from the University of Florida.
$300K+ in credits and growing across 25+ partners. We hustle to get our founders the tools, infrastructure, and services they need to build fast.
Illustrative only. Dilution: seed ~20%, A ~20%, B ~15%, C ~12%. SAFE converts at ~$20M. Carry is 20% above 1x. See full deck appendix for detailed scenarios.