Careers
Small team. Unusually specific problems.
We hire slowly, and mostly for people who are comfortable at the seam between physics and commerce — colour scientists who care about return rates, and engineers who will read a technical report before writing the function.
Where things stand
There are no open listings today.
We would rather say that plainly than keep a page of aspirational roles we are not funded to fill. ExactHue is pre-seed, founded in 2025, and the team is small enough that every hire changes what the company is.
That does not mean don't write. Most of the people we have talked to seriously arrived before there was a role, and a good letter has a long shelf life here. If the problem described on this site is one you have actually thought about, we want to read what you think.
What we are not looking for is a CV with a covering note that could have been sent to fifty companies. Tell us what you would build first and why, or what you think we have got wrong. Two paragraphs of that beats two pages of anything else.
Areas we are usually interested in
- Colour science and appearance modelling
- Computer vision on uncontrolled consumer captures
- Commerce platform integration engineering
- Research operations and participant recruitment
- Data governance and privacy engineering
Practicalities
| Base | Singapore |
| Stage | Pre-seed |
| Team | Small, and honest about it |
| Lab access | On site — the measurement work is physical |
How we hire
Four conversations, no homework marathon.
We think take-home exercises that eat a weekend are a tax on people with other commitments, so we do not set them.
You write, we reply
Every message gets a real answer from a person, including the ones where the answer is no. If we are not hiring for what you do, we will say that rather than leaving you to guess from silence.
A conversation about the problem
Forty-five minutes on what we are building and what you would do differently. This is the part that matters most. Come with opinions.
Working session
Two hours, paid, on something close to real work — reading our benchmark methodology and picking holes in it, or walking through code together. No trick questions and nothing you would not do on a normal Tuesday.
Meet the rest of us
Because at this size you will work with everyone, and you should get to interview us too. Ask about runway, ask about the parts that are not working. You will get a straight answer.
Before you write
Things that are true about working here.
Some of these are selling points and some are not. Both are useful to know in advance.
The measurement work is physical
There is a laboratory, there are calibration targets, and there are days spent measuring samples. If you were hoping for a purely remote research role, this is not that. Some of the most useful insights here have come from someone noticing something odd on a bench.
We ship slower than we could
A sampling plan before a training run, a laboratory before a marketing team. It is a deliberately unfashionable order of operations and it means fewer visible wins in the first year. If that would frustrate you, it genuinely would frustrate you.
Being wrong out loud is normal
The product declines to answer when it is not confident, and the culture is much the same. Saying "I do not know yet" in a meeting is unremarkable here. Saying it after you have shipped the guess is not.
Pre-seed means pre-seed
Runway is finite, the salary will not beat a large company's, and there is real risk. We would rather you weigh that with accurate numbers than find out later. Ask, and we will tell you where we actually are.
Still interested?
Tell us what you would build first, or what you think we have got wrong on this site. Both are good openings.