Back to blog

Hard-Tech Due Diligence: However Compelling the Team Story, Return to Public Data

In my years of investing in hard tech, I have developed a stubborn habit: the more moving the team story, the more rigorous the verification. The logic is not complicated. For hard-tech projects, the valuation anchor is to a large extent “people”—the core scientists’ track records, technical accumul

AI Scientific Research AgentAI Office Assistant AgentAI Agent

In my years investing in hard tech, I have developed a stubborn habit:The more moving the team story, the more detailed the verification.

The logic is not complicated. The valuation anchor for hard tech projects is, to a large extent, “people”—the core scientists’ track records, technical accumulation, and academic networks. And these are exactly the hardest parts to verify in due diligence: the technical narrative is too specialized, making it hard for outside investors to judge the quality of “world-leading” claims; founders’ ability to package the team is often stronger than their technology.

Our current approach is to build a background-check checklist for each member of the core scientist team, using MuchSci’s talent profiles as the source document.

Checklist item one: public track record. Education and work experience are checked paragraph by paragraph against the story the team tells. “Core members come from a top team”—which team, what role, what period; corresponding records should be findable in the public track record.

Checklist item two: project experience. If the team says it “led a national-level project,” check it in the profile. The key point: project details in the profile have undergone dual verification of “exact match of the project leader’s name + validation of the publicly listed institution”—whether the source project behind the technical story actually exists. This item is the most telling.

Checklist item three: academic collaboration network. A public collaboration trail with an evidence chain answers an important question: whether this scientist’s academic support is a real community or a lone effort. Teams with real networks usually have more credible execution capability.

Once these three are verified, we move to cross-checking:

For every claim by the founding team, find the corresponding record in the public profile. Consistent ones are marked “Verified”; discrepancies are flagged separately and the team is asked to explain—former names and institution renamings often have reasonable explanations, but the explanations must come from them proactively, not be smoothed over by us on their behalf.

The style of the due diligence report is thus set:Do not write “it is said.” Every conclusion is followed by a source; if something was not found, say clearly that it was not found.

Back to articles