BioInvent: Practice for High-Stakes Science Teams

BioInvent International is an immuno-oncology company. Its proprietary F.I.R.S.T platform screens for antibodies and the tumor targets they bind at the same time, surfacing immunomodulatory candidates that move into a clinical pipeline. Work like that runs on cross-functional teams: discovery scientists, translational researchers, clinical and project leads, all pushing a complex program forward under real time pressure.

In an environment like that, the conversations that decide how well a team works are not the routine ones. They are the hard ones. Raising a dissenting view in a room full of senior people. Telling a colleague their analysis does not hold up. Aligning two disciplines that see the same data differently, with a deadline closing in. Those moments are where psychological safety is either built or quietly eroded, and most people get exactly zero chances to practice them before they happen for real.

Why do high-stakes science teams need to practice conversations at all?

Research teams rehearse almost everything else. They pressure-test assays, repeat experiments, and review protocols until the method is sound. The interpersonal side of the work rarely gets the same treatment. People are expected to deliver a difficult message well on the first try, in the actual meeting, with real stakes on the table. When it goes badly, the cost is not abstract: a concern goes unspoken, a decision gets made on incomplete pushback, and the next person hesitates a little longer before speaking up.

Psychological safety is the term researchers use for a team where people can take interpersonal risks, voicing doubt, admitting a mistake, asking a naive question, without fear of being punished or dismissed. It is not a personality trait of nice people. It is a property of how a team handles its hardest exchanges, repeatedly, over time. That makes it something you can practice toward, not just hope for.

How does PokaMind fit into BioInvent's work?

PokaMind gives these teams a place to rehearse the specific conversations before they happen for real. An AI avatar plays the other person in the room: the senior colleague hearing a dissent, the teammate receiving hard feedback, the counterpart from another discipline who is not yet aligned. The person practicing has the conversation out loud, as many times as they want, with no one watching and nothing riding on it.

The scenarios are not generic. They are drawn from BioInvent's own context, so a session sounds like a conversation that could actually occur on one of their programs rather than a stock business-school case. During each roleplay, PokaMind reads tone of voice, word choice, facial expression, body language, and pace, then ends the session with specific feedback on how the message landed and what to adjust. The AI is a roleplay partner the team practices against, not a coach handing down advice.

• Raising a dissenting view to a more senior colleague without softening it into nothing.
• Giving direct, useful feedback on work that does not meet the bar.
• Aligning across disciplines when two experts read the same result differently.
• Surfacing a concern early instead of staying quiet and hoping someone else does.

What does practice change in the moment that matters?

The value of a rep is not that the script comes out word for word later. It is that the conversation stops being unfamiliar. Someone who has already raised a dissent three times against a roleplay partner has felt the discomfort, found their footing, and heard how their tone reads when they are nervous. When the real moment arrives, they are working from experience instead of improvising under pressure. That is the difference between a concern that gets voiced clearly and one that gets swallowed.

Where the partnership stands

This is an early partnership, and we want to be honest about that. The premise is straightforward: give high-stakes teams a safe place to practice the conversations that build psychological safety, on their own scenarios, and let the reps do the work. We are not going to dress up outcome numbers or put words in anyone's mouth while the work is still underway. What we can say plainly is that the approach is practice-first, the scenarios come from the team's real context, and the feedback is specific to how each conversation actually lands.

If your teams carry the same kind of high-stakes, cross-functional conversations, whether in research, clinical work, or any setting where speaking up well actually matters, the same model applies. People practice the hard exchange before it counts, on scenarios that look like their own work, and walk in steadier for it.