Venture Models

What young companies might be worth

Independent practice models I built using only public information, each asking what a startup might be worth today and what an investment in it could return.

Every input in the models is tagged VERIFIED (from a cited public source) or ASSUMPTION (my own estimate). This is practice analysis, not investment advice.

Defense testing01

Valkryn

Valkryn runs independent “Battle Lab” tests of military drones, ground robots, uncrewed boats and counter-drone systems, and it does not build the systems it tests.

Round
$15M seed
Lead
a16z American Dynamism
Date
Nov 17, 2026

How the model works

Valkryn hasn’t said what it’s worth, so I estimated its value after this round at about $75 million (somewhere between $60 million and $100 million), based on how much of a company investors usually buy in a round this size and on what similar startups were valued at. Then I asked what a $1 million check at that price could turn into, after later funding rounds shrink its share. A bad outcome returns about 0.2x the money, a normal one about 5.6x and a great one about 22.5x, or about 4.3x on average once I weight each outcome by how likely I think it is. The valuation, exit values and odds are my estimates, because none of them are public.

What a $1M check could become

Bad
0.2x
Normal
5.6x
Great
22.5x
Weighted
4.3x

Estimated value after the round: $75M (range $60M to $100M). All estimates, not reported figures.

Download the model (.xlsx) ↓Download the brief (PDF) ↓

Defense AI software02

Lumbra

Lumbra builds LumbraOS, software that runs AI agents on top-secret defense and intelligence networks and keeps a record of everything each agent does.

Round
$25M Series A
Lead
Point72 Ventures and Ansa Capital (co-leads)
Date
Nov 10, 2026

How the model works

Lumbra hasn’t said what it’s worth, so I estimated its value after this round at about $125 million (somewhere between $100 million and $167 million), using how much of a company investors usually buy at this stage and similar deals, like NODA AI raising the same $25 million at a $125 million valuation. Then I asked what a $1 million check at that price could turn into, after later funding rounds shrink its share. A bad outcome returns about 0.3x the money, a normal one about 5.6x and a great one about 22.5x, or about 4.4x on average once I weight each outcome by how likely I think it is. The valuation, exit values and odds are my estimates, because none of them are public.

What a $1M check could become

Bad
0.3x
Normal
5.6x
Great
22.5x
Weighted
4.4x

Estimated value after the round: $125M (range $100M to $167M). All estimates, not reported figures.

Download the model (.xlsx) ↓Download the brief (PDF) ↓

Space sensors03

Scout Space

Scout Space, based in Reston, Virginia, builds sensors and software that let satellites see and track other objects in orbit.

Round
$18M Series A
Lead
Washington Harbour Partners
Date
First close May 6, 2026

How the model works

Scout Space hasn’t said what it’s worth, so I estimated its value after this round at about $90 million (somewhere between $72 million and $120 million), based on how much of a company investors usually buy in a round this size and on what similar space startups were valued at. Then I asked what a $1 million check at that price could turn into, after later funding rounds shrink its share. A bad outcome returns about 0.2x the money, a normal one about 3.9x and a great one about 17.2x, or about 3.2x on average once I weight each outcome by how likely I think it is. The valuation, exit values and odds are my estimates, because none of them are public.

What a $1M check could become

Bad
0.2x
Normal
3.9x
Great
17.2x
Weighted
3.2x

Estimated value after the round: $90M (range $72M to $120M). All estimates, not reported figures.

Download the model (.xlsx) ↓Download the brief (PDF) ↓