Nexstrom raised $12 million to take a lab result to a 12-inch wafer. The question that tests whether you can find the hard part.
On September 22, 2026, TechCrunch reported that Nexstrom, a Singapore-based semiconductor startup incubated by Xora Innovation, had raised a $12 million seed round from investors including Xora, Foothill Ventures and SEEDS, and has raised $15 million in total. The company is building equipment and processes to grow transition-metal dichalcogenides, materials only a few atomic layers thick, directly on the 300mm wafers used in commercial chipmaking. Its co-founder and chief scientist Lance Li, who previously led research on post-silicon electronics at TSMC, described the problem plainly. "In the lab, people can get very small, high-quality materials, but only on a very small scale," Li told TechCrunch. "This is exactly the gap Nexstrom is addressing." CEO Phoebe Tan told TechCrunch the company has "systematically scaled from 2 inches to 6 inches," with larger wafers targeted for the end of October. TechCrunch noted that some companies in the industry are testing samples, though Nexstrom declined to name them, and that the startup expects its equipment to be ready for commercial production between 2030 and 2035.
Notice what Nexstrom did not claim. It did not say it had solved 2D semiconductors. TechCrunch listed TSMC, Intel, IMEC, AIXTRON and CDimension among the companies working on other parts of the same chain, and wrote that "there are several problems to be solved" before these materials can be manufactured at scale. Nexstrom picked one of them, the gap between a lab sample and a full wafer, and then walked it in steps that could each be measured. That is a way of thinking, not a fact about chips. Take an enormous, well-known problem, refuse to solve all of it, find the specific gap where the difficulty actually lives, and attack that gap in an order that can be checked. Interviewers test for exactly this habit with a question that has nothing to do with wafers: "How would you solve traffic congestion in Seattle?"
Why they ask it
Nobody expects a transport plan. The question is chosen because it is too large to answer, and the interviewer wants to watch what a candidate does with a problem that does not fit in the time available. Does the candidate ask what congestion means and for whom? Do they pick a target they could measure? Do they choose one lever and explain why that one rather than list eight? At senior levels the question is also about honesty on timelines. Nexstrom told a reporter its equipment would be commercially ready between 2030 and 2035, on the same day it announced its funding. A candidate who says a city's traffic can be fixed in a quarter has told the interviewer how they will estimate a roadmap.
The trap
The common failure is the brainstorm. Congestion pricing, more buses, light rail extensions, remote-work incentives, staggered school hours, better signal timing. Each idea is reasonable and the list as a whole is worthless, because it shows no judgment about which one matters. The second failure is the opposite. The candidate picks the first idea they like, usually the most technical one, and defends it without ever defining the problem. Both fail the same way. Neither answer has found the gap. One candidate has described the whole chain and pretended to own it. The other has grabbed a link and pretended it is the chain.
Applying STAR-T
This is a hypothetical, so the frame is used differently. Situation and Task are what the candidate builds in the first minute by narrowing the question. Action is the plan, in sequence. Result is what would be measured. Trade-off is the cost named before anyone asks.
Situation. Establish what congestion is and where. "Congestion in Seattle is not one thing. I would first separate peak-hour commuter delay on the freeways and bridges from downtown surface-street delay, because they have different causes and different owners. I would ask which one the city is being asked to fix, and absent an answer I would assume peak commuter delay, since that is where most of the lost hours are."
Task. Turn that into something measurable. "I would define the target as reducing average peak travel time on the three worst corridors against last year's baseline, rather than solving congestion, which no city has done."
Action. Pick a lever, sequence it, and show what a small result proves and does not prove. "I would start with demand rather than capacity, because adding lanes takes a decade and induces its own traffic. Phase one is a time-of-day pricing pilot on one corridor, with transit frequency doubled on the same route so nobody is priced out without an alternative. Phase two extends to the other two corridors only if delay drops on the first. Phase three uses that data to decide whether capital projects are still needed." The point of the phases is the point Nexstrom made about wafers. A result at small scale is evidence, not proof, and the plan should say what changes at each step.
Result. State what would be examined and when. "At six months I would look at travel time on the pilot corridor, ridership on the parallel transit route, and whether traffic moved to neighbouring streets. If it moved, the pilot failed even if the corridor itself looks better."
Trade-off. Name the cost. "Pricing is regressive unless the transit alternative exists on day one, and it makes the city a political target. I would still choose it over capacity because it can be reversed if it fails, and a new lane cannot."
The follow-up that breaks weak answers
"What would you do if the pilot showed no change?" The brainstormer has nowhere to go and offers the next item on the list. The candidate who grabbed one idea defends the pilot. A strong answer treats a null result as information about the gap. It states what the pilot was meant to test, which assumption the result disproved, and which branch of the plan that closes. Nexstrom's CEO described work "focused on delivering material quality that [industry partners] have specified", which is another way of saying the customer defines pass and fail before the test runs. The strong candidate says who defines success for the pilot, and what happens when it does not pass.
Score your answer against the director’s bar
Q: How would you solve traffic congestion in Seattle?
Pick one open-ended problem from your own domain, narrow it to a gap you could measure, and rehearse the plan until the phases come out in order. Try it free →