Resources

What Is Deep Tech? A Five-Minute Introduction

A clear introduction to the science and engineering behind deep tech, the problems it can solve, and what it takes to create commercial value.

Every business runs into problems that existing tools can’t easily solve: a process consumes too much energy, a factory struggles to fill demanding roles, or a computing system can’t keep up with the work asked of it. Sometimes the answer begins with a scientific or engineering advance.

Deep tech is a term for companies whose competitive advantage rests on a meaningful advance in science or engineering. That advance might enable a machine to handle work in changing environments, a material to perform differently, or software to make computing resources more useful. It can be a new discovery or a difficult new application of established science. The technical work matters because of what it can make possible for customers.

These companies operate in familiar markets, including manufacturing, construction, energy, logistics, and computing. What distinguishes them is the technical advance at the heart of the business—and the opportunity to turn it into something people will pay to use.

What does deep tech look like?

Robotics. Warehouses and factories have long used machines for repetitive tasks. A harder challenge is making robots useful in settings designed for people. Apptronik, a BlueLake.vc portfolio company, is developing human-shaped robots that combine mechanical engineering with software trained on real-world tasks. Mercedes-Benz has tested its Apollo robots in a production environment, exploring uses such as bringing components to a production line. The potential value is more flexible support for physically demanding work and hard-to-fill roles. These efforts remain in testing and development. [1]

Materials. Concrete producers must meet demanding strength requirements while managing material costs. CarbonCure injects captured carbon dioxide into fresh concrete, where it reacts to form a mineral. In some mixes, this allows producers to use less cement while meeting strength requirements. Thomas Concrete has described using the technology at its Riverside plant. [6] Depending on the mix, customers may benefit from lower material costs and a lower-emissions product. [2]

Energy. Utilities need power when demand is high, including when wind or solar output is low. Form Energy is developing an iron-air battery that stores electricity using a reversible version of rusting: iron reacts with oxygen as the battery discharges, and the process reverses when it charges. The company reports that it deployed its first system for commercial demonstration with Great River Energy in 2025. [5] Its goal is longer-duration storage that could help utilities supply electricity through extended periods of grid stress; performance and economics at scale still matter. [3]

Computing. Running AI can require substantial computing capacity, and software built for one type of chip may need considerable work to run on another. Lemurian Labs, another BlueLake.vc portfolio company, is developing Tachyon: software designed to translate and coordinate AI programs across different processors. If successful, it could give customers more hardware choices and reduce the work of adapting their applications. The company describes a compiler and runtime that coordinate computation and move data across a system. Its claims about portability and performance need to be assessed for the customer’s actual workload. [4]

It can be software—or hardware

Lemurian illustrates why deep tech is not limited to physical inventions. A new approach to computing or an algorithm—a set of instructions for solving a problem—can represent a substantial engineering advance.

At the same time, adding AI to a product does not automatically make the company deep tech. Neither does building a physical device from readily available components. Both can make for valuable businesses. The question is whether a meaningful technical advance is central to what the company does and helps give it a competitive advantage.

From breakthrough to business

A promising result in a lab is a starting point. For it to become commercially valuable, a company must turn it into a product that works reliably, solves an important problem, and can be delivered at a price customers will pay.

Consider a hypothetical manufacturing material. A customer might like how a sample performs but still need evidence that it can be produced consistently, meet required standards, work with existing equipment, and arrive when needed. The startup must deliver all of that while earning enough to build a durable business.

Patents and specialized know-how can help protect an invention, but they do not guarantee customer adoption or profits. Even a successful company can be a disappointing investment if the price paid, later fundraising that reduces an investor’s ownership, or the time before a sale or public listing leaves too little return.

What investors need to understand

Deep tech investments can involve several kinds of uncertainty:

  • Technical: Does the technology work reliably outside a controlled test, and what remains unproven?
  • Operational: Can the company produce and deliver it at the required quality, volume, and cost?
  • Commercial: Who will buy it, how much value does it create, and how difficult is it to replace the customer’s current solution?
  • Regulatory and timing: Are permits, certifications, or other approvals needed, and how long could they take?
  • Financial: How much capital will the company need to reach its next milestones, and what could future fundraising mean for existing investors?

These uncertainties can move at different speeds. A company might have strong technical results without paying customers, or early customers while still working to make each product profitably. The task is to understand what is known, what remains to be proven, and how much time and capital it may take to bridge the gap.

Deep tech offers a way to think about progress in practical terms: better tools for consequential problems. The opportunity lies in turning those tools into products that customers value and businesses that can endure.

Sources & further reading

  1. Mercedes-Benz: Testing Apollo in production
  2. CarbonCure: How the technology works
  3. Form Energy: Iron-air battery technology
  4. Lemurian Labs: Tachyon technology
  5. Form Energy: Company-reported demonstration milestones
  6. Thomas Concrete: Commercial use of CarbonCure
  7. BlueLake.vc: Selected portfolio

Company examples use public information. Company-reported capabilities and plans may change. Apptronik and Lemurian Labs are selected BlueLake investments; other companies mentioned are illustrative examples.

For educational purposes. This material is not investment advice or an offer to invest.