For over 80 years, America’s science fair system has been quietly selecting the nation’s most capable young builders — teenagers who choose a hard, unsolved problem and spend a year pushing the boundary of what’s known. From Westinghouse (1942) to Intel to Regeneron, these competitions have functioned as the country’s deepest talent audit: deep technical ability, the obsession to ship, and the drive to stand up and defend the work.
Science Fair Fund is the first venture fund built entirely around this community — identifying alumni founders before institutional capital arrives, scoring them with proprietary data no one else has, and concentrating capital behind the ones the model says will break out. This isn’t a thesis about sectors or stages. It’s a thesis about people — that the teenagers who chose to do something hard when no one was watching become the founders who do it again when everything is on the line.
71% graduate to Series A · signal strengthens · AUC 0.84
Series A Follow-Ons · 18 Companies · $8M Reserve (via oversubscription or SPV)
Q1Super Pro Rata60% reach $500M+
6
Concentrated Follow-On
Expand beyond pro rata · 2.15x lift vs. cohort · highest conviction
Q2Pro Rata28% reach $500M+
12
Maintain Position
Pro rata protects ownership in cohort winners
How Scoring Drives Portfolio Construction
The ML model scores every alumni-founded company at each stage of its lifecycle. At founding, competition results, education trajectory, and sector signals identify the highest-potential founders before institutional capital enters.
At first funding, the model augments biographical signals with round size, investor quality, and co-founder composition to size initial checks.
Quartile assignment directly determines capital allocation: Q1 companies receive 2x the check size and get super pro rata follow-on at Series A. This scoring-driven concentration is the mechanism that converts the alumni network’s structural advantage into portfolio returns.
The Structural Moat
Edge
What It Means
Community Flywheel
Participating in science fairs is a formative experience (age 14–18) — the kind that bonds people for life. That bond is the moat. 2,213 alumni said “willing to help” (70%+ response rate). The network compounds with every cycle: alumni open doors at their companies, write checks into each other’s rounds, and make the introductions that move companies forward. Active portfolio founders become the fund’s best source of referrals into the next generation of alumni startups.
Insider Information Asymmetry
ISEF 1st Place (’04) + multi-year Grand Awards Judge, with a proprietary dataset covering 25 years of ISEF alumni and STS records back to 1943 — 49,183+ profiles, the most comprehensive map of this talent pool anywhere. An ML model (55 features, walk-forward AUC 0.84) identifies high-potential alumni before they found companies. Automated monitoring flags career transitions and founding signals in real time — first-ticket SAFEs before institutional VCs see the deck. Relationships with past alumni founders surface repeat founders and warm referrals before rounds are announced.
Proven Alpha
151 alumni-linked companies tracked (2006–2020). At founding — before any institutional capital — Q1 (top quartile, n=38) hits 52.6% $100M+ outcomes (2.6× baseline) and 28.9% unicorn rate (2.6×). Signal strengthens through the lifecycle. Q4 (bottom quartile): 0% across the board at every stage.
The Manager: Anthony Atlas
I’m a Science Fair Winner. I won 1st at ISEF in 2004. That competition changed the trajectory of my life — and I’ve spent 20 years watching it do the same for others. I’ve been a Grand Awards Judge for multiple years and have deep relationships across the alumni community. This isn’t a thesis I researched — it’s a network I grew up in.
I Built the Database. Since 2006, I’ve built the most comprehensive dataset on this talent pool anywhere — 49,183+ profiles, proprietary scoring, and a 20-year backtest. No one else has this data.
I Can Help Them Win. Raised >$75M in venture capital as an operator. Supported 3 deep-tech companies through ~10× valuation step-ups (seed → Series B). I know what early founders need because I’ve been in the room.
How We Compare
Metric
SFF Alumni
Benchmark
Multiple
Unicorn rate
14%
~4.6% 1
3.1x
Series A graduation
71%
~50% 2
1.4x
Loss rate
~9%
30–40% 3
0.3x
Scoring AUC
0.84
N/A
—
1 YC unicorn rate (Hendricks & Howell, 2024; PitchBook 2023 YC cohort analysis).
2 Industry seed-to-Series A rate (Carta, 2024).
3 VC industry total loss rate (Correlation Ventures; Cambridge Associates).
Fund I focuses exclusively on ISEF and STS alumni. The scoring framework is designed to be extensible to analogous communities — Hertz Fellows, RSI, Math Olympiad winners, European science competitions — where formative competition selects for the same founder traits.
Get the Fund Deck
Full investment thesis, team background, and portfolio construction in a 15-page deck.
Current Pipeline
Early-stage alumni-founded companies scoring in the top half (score ≥ 50), founded 2021+. Anonymized for LP review.
—
Pipeline Companies
—
Capital to Deploy
—
Avg P($100M+)
—
Industries
ISEFSTSBoth
Portfolio Outcome Projections
Using calibrated per-company probabilities from the ML scoring model, applied to the current pipeline as a hypothetical portfolio.
—
Expected $100M+
—
Expected Unicorns
Outcome Range (P25 – P75)
Methodology: Each company’s outcome probability is treated as an independent Bernoulli trial. Aggregate distributions computed via Poisson approximation. Range shows 25th–75th percentile; marker indicates median.
Monte Carlo Simulation
2,000 simulations · SFF scored · 127 real outcomes
Fund Model Controls
$—M
— first checks + — follow-on + 20% fees
Portfolio Construction (30 pipeline companies)
Q1 Companies ⓘ
12
Q2 Companies
18
Q1 Check Size
$200K
Q2 Check Size
$100K
Follow-on Reserve ⓘ
$8.0M
Scoring Fidelity ⓘ
80%
Entry & Dilution Assumptions
Seed Post-Money ($M) ⓘ
$10M
Dilution per Round ⓘ
20%
Unrealized Positions
Valuation Mode
Liquidity Discount
40%
Scenario Options
Exit Value Cap
Exclude Top Unicorns
— companies · $—M first checks · $—M follow-on · $—M fund
Gross MOIC Distribution
Scenario
Gross MOIC
Net TVPI
Unicorns
$100M+ Exits
Gross Proceeds
P25 (Conservative)
-
-
-
-
-
P50 (Median)
-
-
-
-
-
P75 (Upside)
-
-
-
-
-
Return Probabilities
≥5x
-
≥10x
-
≥25x
-
≥50x
-
≥100x
-
Link copied to clipboard!
Methodology: Bootstrap from Real Outcomes
Each simulated portfolio draws company outcomes from a pool of 127 real alumni-founded companies (funded, VC-investable, founded ≤ 2017). Two sampling modes are available:
SFF Scored uses the ML model’s predictive power to weight draws — Q1 samples with 2.2× weight on $100M+ outcomes (AUC 0.74 at funding, walk-forward CV);
Random draws uniformly from the pool, representing what any seed investor would get without the scoring edge.
Outcome Pool (127 companies, founded ≤ 2017)
Outcome Tier
Count
% of Pool
Example Values
Source
$1B+ (Unicorn)
21
16.5%
$1B – $500B
Database
$100M – $999M
16
12.6%
$100M – $700M
Database
$1M – $99M
10
7.9%
$5M – $75M
Database
No known valuation
80
63.0%
45 active (no public valuation) · 19 acquired (value unknown) · 8 closed · 7 unknown · 1 IPO micro-cap.
Tier Selection — SFF Scored Mode
Quartile
Scoring Range
$100M+ Rate
$1B+ Rate
Source
Q1 (Gold)
Top Quartile (n=33)
62.5%
34.4%
Database
Q2 (Teal)
Second Quartile (n=33)
36.4%
18.2%
Database
Q3
Third Quartile (n=33)
9.1%
9.1%
Database
Q4
Bottom Quartile (n=33)
0%
0%
Database
Pool Baseline (Random mode)
All (n=127)
29.1%
16.5%
Database
Scoring Fidelity — Model Accuracy Discount
Backtest tier rates assume perfect classification — every company the model ranks as Q1 truly belongs there. In practice, no model is perfect. Scoring Fidelity controls what fraction of companies are drawn from their assigned quartile’s outcome distribution vs. the overall base-rate distribution.
At 100% fidelity, every company draws from its scored tier (full backtest rates).
At 80% (default, calibrated to walk-forward AUC = 0.74 at first funding), each company has a 20% chance of drawing from the base pool instead — modeling the probability that the scoring model mis-ranked it.
At 50%, tier assignment is essentially random and returns converge to an unscored portfolio.
Fidelity
Interpretation
Effective T1 $100M+ Rate
100%
Perfect classifier (full backtest)
62.5%
80% (default)
Calibrated to AUC at first funding
55.8%
50%
Random classification (no signal)
45.7%
Dilution Model (round-count based)
Parameter
Value
Notes
Source
Entry stage
Seed
Post-money: $5M–$25M (adjustable, default $13M)
Adjustable
Per-round dilution
20%
Each round after entry (adjustable, 15%–30%)
Industry
Avg rounds: unicorns
4.7
From actual company data
Database n=20
Follow-on (T1)
Super pro rata
Offsets 2 dilution rounds
Industry
Follow-on (T2)
Pro rata
Offsets 1 dilution round
Industry
Fund Economics (Fixed)
Parameter
Value
Notes
Source
Management fee
2%
× fund size × 10yr life (modeled as 20% upfront)
Industry
Carry
20%
Of profits above 1× return of capital
Industry
Scoring & Backtest
Scoring System, Evaluation Framework & Results
Executive Summary
Science-fair competitions (ISEF, STS) select for a rare combination of technical depth, independent research ability, and competitive drive — traits that compound in venture-backed founders. Science Fair Fund uses a proprietary ML scoring system to rank these alumni at each investment stage, enabling three portfolio construction decisions: prioritizing sourcing, sizing initial checks, and concentrating follow-on capital.
The scoring system’s edge is early identification: at founding — before any institutional capital — the top quartile produces $100M+ outcomes at 2.60× the baseline rate. Signal strengthens through funding and Series A as information accumulates, enabling disciplined capital allocation at every stage.
240
Companies Tracked
105
Series A+
0.84
CV AUC (Series A)
1.80×
Q1 Lift (Series A)
Primary metric: $500M+ outcomes (stable sample, directly relevant to fund returns). $100M+ and $1B+ shown for context.
Two evaluation frameworks: Selection Advantage (tier separation within each stage) and Model Validation (walk-forward CV + time-gated backtests).
At Series A: Q1 concentrates 50% $500M+ hit rate (1.80× lift).
Data Universe
Reference cohort = VC-investable alumni-linked companies founded 2006–2020, with outcomes observed through early 2026. Every company has at least five years of maturity.
Metric
Full Cohort
Funded ≥$500K
Series A+
Alumni-linked companies
240
148
105
$100M+ outcomes
38 (15.8%)
37 (25.0%)
37 (35.2%)
$500M+ outcomes
25 (10.4%)
25 (16.9%)
25 (23.8%)
$1B+ outcomes
21 (8.8%)
21 (14.2%)
21 (20.0%)
172 of 240 companies (71.7%) are US/Canada-headquartered, consistent with the fund’s North American deployment focus.
Scoring Methodology
The fund operates two complementary evaluation systems, each answering a different question about alumni founders.
Ranking Model (V4 Ensemble). Each company receives a stage-specific score based on its highest-scoring founder. The score is a within-cohort rank — not a calibrated probability — designed to sort companies relative to peers at each decision point. It answers: who should we prioritize?
Outcome Prediction (Walk-Forward CV). Separately, a walk-forward cross-validated model tests whether the ranking generalizes forward in time — trained on earlier cohorts, evaluated on later ones — using only information available at each stage. It answers: does the model actually predict who wins?
Three Decision Points
At Founding. Competition results, education trajectory, and sector signals available before institutional funding. Focuses sourcing on the highest-potential alumni.
At First Funding. Founding signals augmented by round size, investor quality, and co-founder composition. Primary signal for initial check sizing.
At Series A. Funding signals plus milestone progression — time to Series A, round scaling, traction markers. Drives follow-on concentration.
Signal Evolution by Stage
How signal categories shift in importance as companies mature — the model adapts its weighting at each decision point.
Signal Category
At Founding
At Funding
At Series A
Competition Record
●●●
●●○
●○○
Sector Fit
●●●
●○○
●●○
Capital Progression
—
●●●
●●●
Technical Depth
●●○
●●○
●●●
●●● = high ●●○ = medium ●○○ = low — = not available at this stage
Model Validation
The predictive test: does the model generalize forward in time? Walk-forward cross-validation trains on earlier founding-year cohorts, evaluates on later ones, and uses only information available at each stage.
Decision Point
Walk-Forward AUC
Fund Application
At Founding
0.67
Sourcing priority within alumni universe
At First Funding
0.74
Initial check sizing ($100K–$250K range)
At Series A
0.84
Follow-on concentration decisions
An AUC of 0.84 at Series A means the model correctly ranks a random $100M+ outcome above a random non-outcome 84% of the time.
Precision/Recall at Q1 Threshold
Stage
Selection %
Precision ($500M+)
Recall ($500M+)
At Founding
25%
34.2%
61.9%
At First Funding
26%
37.5%
50.0%
At Series A
25%
50.0%
45.5%
Selection Advantage
The ranking model’s cross-sectional test: within each stage cohort, does the score concentrate future winners into a smaller, actionable subset?
Across all three decision points, the scoring system concentrates $500M+ outcomes toward the top of the ranked list.
Stage cohort sizes differ because each decision point uses its own eligible universe (151 at founding → 125 funded → 79 Series A+).
Early signal is observable before institutional capital. The founding score shows Q1 $100M+ rates at 2.60× the baseline rate.
Signal strengthens with information. Q1 $500M+ lift ranges from 1.80–2.50× across stages.
The score is a rank, not P($500M+). The V4 score is a percentile rank. $500M+ probabilities are separately calibrated.
Definitions
$100M+
Exit value or last known valuation ≥$100M.
$500M+
Exit value or last known valuation ≥$500M.
$1B+ (Unicorn)
Exit value or last known valuation ≥$1B.
Funded
Institutional capital raised ≥$500K.
Series A+
Venture stage at Series A or later, funded ≥$500K.
Walk-Forward CV
Trains on earlier cohorts, tests on later ones. No future data leaks into training.
AUC
Area Under ROC Curve. Probability a positive is ranked above a negative. 0.5 = random; 1.0 = perfect.
Precision
Fraction of selected companies that are actual $500M+ outcomes.
Recall
Fraction of all $500M+ outcomes captured by the selected set.
Lift
Outcome rate in a tier divided by the baseline rate for the full stage cohort.
SHAP
SHapley Additive exPlanations. A method that assigns each feature a contribution score for a given prediction, based on cooperative game theory.
Alumni Founded Companies
Top 50 alumni-founded companies with $100M+ outcomes
178
Funded Companies
—
Unicorns
—
Realized Exits
—
Enterprise Value
ISEFSTSBoth
Facts & Methodology
Everything an investor (or AI) needs to evaluate the thesis.
1. The Thesis
Science fair alumni are the most under-recognized founder talent pool in venture capital. Every year, thousands of teenagers compete at ISEF (International Science and Engineering Fair) and STS (Science Talent Search). These competitions select for a rare combination: deep technical ability, independent research capacity, and the obsession to ship results under pressure — exactly the traits that predict startup success.
Despite producing 14% unicorn founders (3.1x the venture industry baseline of ~5%), this cohort receives almost no structured institutional capital at the moment of founding. Science Fair Fund exists to close that gap — identifying alumni founders before institutional capital arrives and concentrating capital behind the highest-conviction opportunities.
The fund’s structural edge is threefold: (1) proprietary database of 49,183+ alumni that no other investor maintains, (2) an ML scoring model trained on 20 years of alumni outcomes, and (3) tribal access — the GP is an ISEF 1st Place Winner (’04) with deep community relationships.
2. America’s Farm System for Innovation
In 1942, Westinghouse Electric launched the Science Talent Search to find the nation’s most promising young scientists. It wasn’t philanthropy — it was talent infrastructure. The STS became known simply as “The Westinghouse,” a 57-year institution that functioned as an intellectual audit of America’s high schools. Seniors submitted original research papers judged on depth and rigor, often conducted in university labs. Alumni include Ray Kurzweil (1965) and Regeneron co-founder George Yancopoulos (1976). SFF’s database includes STS records back to 1943.
Meanwhile, the broader ecosystem of local and regional science fairs — organized by the same nonprofit, Science Service (now Society for Science) — coalesced into a national competition in 1950 and went international in 1958 as ISEF. Where the Westinghouse was the elite selection, ISEF was the farm system: a pyramid of ~400 affiliated regional fairs worldwide, with roughly 175,000 students competing each year for ~1,700 finalist spots. Projects are presented on display boards and defended in live oral interviews — selecting for the ability to build, communicate, and hold up under scrutiny.
After Sputnik, Congress passed the National Defense Education Act, and overnight these fairs shifted from hobby to national priority. The two programs served complementary functions: STS was the intellectual audit, ISEF was the public spectacle. Students often competed in regional fairs as underclassmen before submitting to STS as seniors. Westinghouse subsidized Science Service — the organization that administered both — effectively funding the pipeline that fed its own prestige competition. Together they created a talent system that has been quietly compounding for over 80 years.
The sponsorship passed from Westinghouse (1942–1998) to Intel (1998–2016/2019) to Regeneron (2016/2020–present) — and that last transition closed a remarkable loop. George Yancopoulos was himself a top Westinghouse STS winner in 1976. He has said the competition was a pivotal moment that gave him the confidence to pursue science. Forty years later, he brought the sponsorship home — the first time a single entity has held title sponsorship of the entire ecosystem.
SFF tracks alumni from both competitions. The fund’s thesis rests on the observation that these programs have consistently selected teenagers with the specific combination of technical depth, independent drive, and competitive resilience that compounds in venture-backed company building. Today this ecosystem produces ~2,000 finalists per year — teenagers who chose to spend a year solving a hard problem when no one was making them. SFF’s database tracks what happens to them next.
3. Database Methodology
The SFF Alumni Intelligence Engine contains 49,183+ verified alumni profiles spanning ISEF and STS competitions from 1950 to present. Data sources include official competition records, Society for Science archives, professional profile enrichment, and public company databases.
Each alumni record includes: competition history (year, placement, project), educational trajectory (undergraduate and graduate institutions), career path (current role, company, location), and founder status (linked to specific companies with founding dates and roles).
Professional profile enrichment covers approximately 17% of the database; for funded founders, coverage is substantially higher. The database identifies 178+ companies with $500K+ in funding and tracks over $750B in aggregate enterprise value across exits and current valuations.
Data quality is maintained through automated health checks, deduplication pipelines, and materialized views that are refreshed after every import. All LP-facing metrics are derived from a standardized definitions file to ensure consistency across reports.
4. Scoring Model
The SFF scoring model is a gradient-boosted classifier trained on the full alumni-founded company universe. The model predicts $100M+ outcomes using 55 features across five categories: competition results, education, company fundamentals, founder biography, and sector signals.
Best AUC: 0.84 at Series A stage. At founding (before any funding data is available), the model still achieves meaningful separation between tiers. Walk-forward cross-validation is used to prevent data leakage — the model is always trained on companies founded before the test period.
Q1 alumni (top quartile by score) show 18% unicorn rate and 37% $100M+ outcome rate — approximately 2x the baseline cohort rates. This lift is consistent across all three scoring stages (at founding, at first funding, at Series A).
The model is retrained quarterly as new outcome data becomes available. Feature importance shifts predictably across stages: competition results dominate at founding, while funding round characteristics become more important at later stages.
5. Backtest Methodology
The Monte Carlo simulation uses a 127-company real outcome pool: 47 companies with known positive outcomes (ranging from $5M to $500B) and 80 companies with zero exit value. This is not synthetic data — every company in the pool is a real alumni-founded venture with verified outcomes.
Simulation methodology: bootstrap resampling from the outcome pool with replacement. Each simulation run constructs a full portfolio of 30 companies (10 Q1 at $200K, 20 Q2 at $100K), applies position-level dilution modeling, and calculates net fund returns including management fees and carry.
Conservative assumptions: (1) 80% fidelity — the model captures only 80% of the historical signal, (2) 40% liquidity discount on unrealized positions, (3) exit caps at $10B to limit tail-driven results, (4) no credit for follow-on allocation alpha.
The interactive simulator on the Fund Returns tab allows investors to adjust all assumptions and see the impact on return distributions in real time.
6. Fund Structure
Fund I: $5M target. Stage: pre-seed and seed. Portfolio: 30 companies deployed over 2 years. Check sizes are quartile-based: $200K for Q1 (top quartile by ML score, 10 companies) and $100K for Q2 (second quartile, 20 companies). Total first-check capital: $4M.
Follow-on reserve: $8M allocated for Series A pro rata rights (via oversubscription or SPV). Q1 companies receive super pro rata follow-on; Q2 companies receive standard pro rata. This reserve strategy concentrates capital in winners identified by both the scoring model and market validation.
Management fee: 2% annually on committed capital during the investment period. Carried interest: 20% above a preferred return. Fund term: 10 years with two 1-year extensions.
7. Comparable Benchmarks
Metric
SFF Alumni
Benchmark
Unicorn rate (funded companies)
14%
~4.6%
Series A graduation rate
71%
~50%
Loss rate (total write-off)
~9%
30–40%
$100M+ outcome rate
25%
~15%
ML scoring AUC (Series A)
0.84
N/A
Sources: SFF Alumni Intelligence Engine (alumni cohort); unicorn benchmark ~4.6% YC (Hendricks & Howell, 2024; PitchBook); Series A rate (Carta, 2024); loss ratio (Correlation Ventures; Cambridge Associates). Cohort: alumni-founded companies with $500K+ funding, 2006–2020.
8. Important Disclosures
For Accredited Investors Only. This material is provided for informational purposes only and does not constitute an offer to sell or a solicitation of an offer to buy any securities. Any such offer would be made only pursuant to a definitive offering memorandum and subscription agreement.
Past performance is not indicative of future results. The backtest results presented are based on historical data and Monte Carlo simulation. Actual fund returns may differ materially from simulated results. All investments involve risk, including the possible loss of principal.
The scoring model’s predictive accuracy (AUC 0.84) is measured on historical data and may not persist in future periods. Model performance may degrade as market conditions change.
Science Fair Fund is not registered as an investment adviser under the Investment Advisers Act of 1940. This communication is not investment advice. Prospective investors should consult their own legal, tax, and financial advisors before making any investment decision.
Welcome!
Two quick questions to personalize your experience.