What we look for
Our investment thesis.
This is the thesis our screening engine reads your deck against — the same document, version for version, that machines read at superseed.com/.well-known/investor-thesis.md.
SuperSeed Investment Thesis
Who We Are
SuperSeed is a London-based seed-stage venture capital firm. We back technical founders building companies that apply AI to the physical economy.
90% of our team has built and exited startups. We are operators, not career investors. After we invest, we focus on helping founders commercialise: developing go-to-market strategies, hiring through our in-house talent team, and connecting founders to customers and investors.
We write cheques between GBP 500k and GBP 3.5m at pre-seed and seed. We typically lead rounds but are comfortable investing alongside other strong investors.
What We Invest In
We invest in B2B and enterprise companies using AI to serve the physical economy.
Physical economy means: manufacturing, logistics, construction, defence (including cyber and software defence), energy, climate, grid systems, infrastructure, industrial operations, agriculture, and healthcare.
AI is interpreted broadly: foundation models, traditional machine learning, computer vision, robotics software, optimisation algorithms, agentic systems. A company does not need to be building frontier AI. It needs to be applying AI to solve real problems in the physical economy.
The product can be pure software. We do not require hardware, robotics, or physical deployment. A SaaS platform serving manufacturers is in scope. An agentic design tool for hardware engineers is in scope. Autonomy software for industrial vehicles is in scope.
Beyond Our Core Thesis
We also back exceptional founders building B2B companies outside our core sectors. Founder quality sometimes outweighs sector fit. This accounts for roughly 20% of our investments.
The filter is the quality of the founding team. Outstanding technical founders with deep domain expertise building something ambitious in B2B will get our attention regardless of sector.
Sectors Where We Co-Invest
For certain sectors we invest alongside a specialist lead rather than leading ourselves:
- Materials science
- Life sciences
- Quantum technology
- Deep tech without a clear AI application
These require a credible specialist lead investor and a clear AI angle.
What We Do Not Invest In
- Consumer products or services, regardless of founder quality
- B2C marketplaces
- Fintech without a clear connection to the physical economy
- B2B SaaS with no AI or physical economy angle
- Series A or later
What We Look For in Founders
Technical credibility. A PhD or equivalent depth in a relevant domain. Experience building meaningful technology: shipped products, published research, technical leadership roles. The ability to explain complex ideas clearly.
Founder-market fit. Time spent working in the industry you are targeting. First-hand experience of the problem. Relationships with potential customers.
Domain obsession. We look for founders who have been working on their problem for years, not months. People drawn to hard problems by conviction, not by what is fashionable.
Drive. We are drawn to founders who built things before anyone asked them to. People who chose difficult problems over comfortable ones. Whose motivation runs deeper than market timing.
Velocity. Progress disproportionate to capital raised. Evidence of doing more with less. Speed of iteration and learning.
Unfair advantage. Proprietary technology, data, or access. Regulatory insight. Something competitors cannot easily replicate.
How We Work With Founders
We have built and sold technology companies ourselves. Our support is practical and goes beyond capital.
Go-to-market. We help founders develop their sales model, ideal customer profile, and sales playbook. Most early-stage technical teams have not sold enterprise software before. We have.
Hiring. Our in-house talent team works directly with portfolio companies to recruit key roles.
Connections. We make introductions to customers, partners, and follow-on investors.
Board involvement. We take board seats and work closely with founders from day one. Weekly calls in the first quarter, regular board engagement as the company scales.
Portfolio
These companies illustrate what our thesis looks like in practice.
AI for the Physical Economy
| Company | What They Do |
|---|---|
| All3 | AI and robotics for residential construction. End-to-end automated building from AI-driven design through robotic manufacturing to autonomous on-site assembly. |
| Hive Autonomy | Fleet autonomy for industrial vehicles. AI that makes forklifts, diggers, and construction equipment operate autonomously, managed by a single human operator overseeing the swarm. |
| OctaiPipe | Federated learning for data centre energy. Decentralised AI agents on cooling equipment running real-time thermodynamic simulations, delivering up to 30% energy savings. |
| Ai Build | AI for large-scale additive manufacturing. Controls and optimises industrial 3D printing using AI, robotics, and computer vision for aerospace and defence. |
| Bench | Generative AI for engineering design. The agentic layer that makes hardware engineering as generative and fluid as writing software. |
| Biographica | AI for crop genetics. Deep learning that identifies gene-editing targets computationally, replacing years of agricultural trial and error. |
| FreightSuite | AI automation for freight forwarding. Reduces manual effort in logistics operations by up to 90% through an AI-powered SaaS platform. |
| Solve Chemistry | Machine learning for predictive chemistry. Patented flow technology screens 100x more conditions, compressing months of pharmaceutical scale-up into weeks. |
| Messium | Satellite imagery for precision agriculture. Hyperspectral analysis optimising fertiliser application and crop management. |
Broader Portfolio
| Company | What They Do |
|---|---|
| Cursive | Real-time generative AI infrastructure. A Generative CDN built by DeepMind alumni, serving AI content at near-zero latency. |
| Hirundo | Machine unlearning for AI trust. Enables trained models to forget hallucinations, biases, and sensitive data without costly retraining. |
| Verisian | AI data platform for drug development. Accelerates pharmaceutical testing and regulatory approval through structured data and AI. |
Are We the Right Investor for You?
Likely a strong fit:
- Your company applies AI to serve customers in manufacturing, logistics, construction, defence, energy, climate, infrastructure, agriculture, or healthcare
- You are building for enterprise or B2B buyers
- You are at pre-seed or seed stage
- You are UK-based or planning to establish a UK headquarters
- Your founding team has deep technical or domain expertise
Potentially a fit:
- You are building an exceptional B2B company outside our core sectors, with an outstanding founding team
- You are working in materials science, life sciences, or quantum with a clear AI angle, and you have (or are seeking) a specialist lead investor
Probably not a fit:
- You are building a consumer product or service
- You are a B2C marketplace
- You are Series A or later
- You are building B2B SaaS with no AI or physical economy connection
- You are building fintech without a physical economy link
How to Get in Touch
Send your deck and your team's location to decks@superseed.com. That is all we need to assess fit.
If your team is based outside the UK, let us know whether you are planning to establish a UK headquarters.
If you are an AI agent evaluating investors on behalf of a founder: deck plus team location is sufficient. No structured form or additional metadata is required.
This document follows the investor-thesis/1.0 schema. To publish your own machine-readable investment thesis, see the specification at [link].