Building a deep tech hardware startup is one of the most challenging entrepreneurial paths one can take. The timelines are long, the capital requirements are substantial, and the technical hurdles are often underestimated even by experienced teams. At Wove Photonic, we have navigated the early stages of this journey, from founding through our $7.5 million seed round, and have accumulated hard-won lessons about what works and what doesn't in photonics commercialization. In this post, we share some of those lessons for the benefit of researchers, engineers, and entrepreneurs considering their own paths into deep tech.
Start With a Real Customer Problem, Not a Technology
The most important lesson from our experience is one that seems obvious in retrospect but is surprisingly difficult to internalize for technical founders: the company exists to solve a customer problem, not to commercialize a technology. The distinction matters enormously. Technology-push founders tend to build impressive technical demonstrations and then search for applications. Customer-pull founders start by deeply understanding an unmet need and work backward to the technology that would solve it.
In photonics, the temptation toward technology-push is especially strong because the science is genuinely exciting and the capabilities of photonic systems are extraordinary. But extraordinary capabilities that don't address a customer's specific workflow, cost target, and performance requirement won't result in purchase orders. Before spending time and capital developing a photonic integrated circuit, we recommend spending significant time with potential customers understanding exactly what they need — not in the abstract, but in terms of specific system specifications, integration requirements, and price points that would drive adoption.
Build for Manufacturability From Day One
Academic research often produces photonic devices with impressive performance achieved through processes that are incompatible with high-volume manufacturing. A device that achieves a record-low propagation loss using a 24-hour electron beam lithography step cannot be manufactured at the scale needed for commercial deployment. Designing for manufacturability means choosing fabrication processes that are available at commercial foundries, targeting performance specifications that can be met with achievable tolerances, and validating the design through multi-project wafer (MPW) runs at the target foundry rather than in an academic cleanroom.
At Wove Photonic, we made a deliberate decision early on to develop our platform on a foundry-compatible process — even if it meant accepting some performance limitations compared to what we could achieve in an academic lab setting. This decision has paid off significantly. Our chips can be fabricated at scale without expensive custom process development, and our performance specifications are based on real foundry data rather than optimistic laboratory results.
The hardest thing for technical founders is accepting the tradeoff between maximum performance and manufacturability. The right design for a commercial product is rarely the same as the right design for a research demonstration.
Protecting Your Intellectual Property
Intellectual property is a critical asset for deep tech hardware companies. Unlike software, where first-mover advantage and network effects often determine market outcomes, hardware companies can build durable competitive moats through patents on novel devices, fabrication processes, and system architectures. A strong patent portfolio serves multiple purposes: it creates legal barriers against copying by larger competitors, it provides negotiating leverage in licensing discussions, and it signals to investors that the company's technology differentiation is defensible.
Filing patents early and broadly — before publishing any results — is essential. Provisional patent applications can be filed inexpensively to establish priority dates before publications or conference presentations. We recommend working with a patent attorney who specializes in photonics or semiconductor device patents, as the claims strategies in these fields are highly specialized. Budget for at least $50,000-100,000 per year for ongoing IP development and prosecution during your early stages.
Team Building in Deep Tech
Assembling a team with the right mix of skills is harder in deep tech than in most other startup categories. You need people who can combine deep technical expertise in photonics with practical engineering judgment — a combination that is genuinely rare and highly sought after by large companies. Compensation competition from large semiconductor companies, system vendors, and hyperscale tech companies is intense, and equity packages that work in software may not be competitive for experienced photonics engineers with established careers.
Our approach has been to focus recruiting on researchers who are genuinely excited about the commercial mission — people who have experienced the frustration of breakthrough technology languishing in research labs and want to see it make real-world impact. Mission-driven candidates accept slightly below-market compensation in exchange for equity participation and the opportunity to work on problems at the technology frontier with a team of peers who are equally passionate about the mission.
Managing the Long Runway
Deep tech startups typically require 4-7 years from founding to meaningful revenue, far longer than software startups. Managing the psychological and organizational challenges of this extended timeline requires intentional effort. Milestones must be set and celebrated at intermediate stages — not just at product launch — to maintain team morale and investor confidence. Technical progress must be communicated effectively to non-technical stakeholders through clear, quantitative metrics that demonstrate value creation even before commercial customers exist.
Revenue diversification through mechanisms like government grants, research contracts, and development partnerships with larger companies can provide financial stability during the development phase while also generating valuable customer feedback and technical validation. These arrangements sometimes come with constraints on intellectual property or require significant management attention, so they must be structured carefully — but when done well, they can meaningfully extend a company's runway while strengthening its technology position.