About ExactHue
A measurement company that happens to work in beauty.
ExactHue Pte. Ltd. was founded in 2025 on a single conviction: that shade matching is not a taste problem to be smoothed over with filters, but a physics problem with a correct answer.
Why we exist
The last thirty centimetres.
Cosmetic chemistry is extraordinarily precise. A foundation shade is formulated to a tolerance that most industries would envy, batch-controlled, and verified with a spectrophotometer before it ever reaches a filling line. The pigment is not the problem.
The problem begins after the product leaves the lab. It is photographed under studio lighting, colour-graded for a campaign, compressed for the web, and rendered on a display nobody profiled — then judged by a customer in a bathroom at nine in the evening. Four uncontrolled transformations sit between a precisely-made product and the decision to buy it, and the industry's response has been to make the preview prettier rather than truer.
The cost lands as returns. Roughly one in four beauty orders comes back, and shade mismatch is consistently among the largest reasons. The cost also lands unevenly: accuracy in most systems degrades toward the deep end of the tonal range, which is exactly where trust was already thinnest and where a bad experience is least likely to be forgiven.
ExactHue exists to close that gap with measurement. Not a better filter. Not a more confident model. A pipeline that models the light, recovers the physical quantity, and reports a number you can check against an instrument.
What we believe
- A number beats an adjective. “Great match” is unfalsifiable. ΔE00 0.7 is not.
- Averages hide the failure. Accuracy reported per tonal band, always.
- Refusing to answer is a feature. A declined capture beats a confident wrong shade.
- Coverage is designed, not discovered. A balanced library is a sampling plan, not a happy accident.
- The preview must survive the parcel. If it flatters, it will be returned.
Company facts
| Legal name | ExactHue Pte. Ltd. |
| Founded | 2025 |
| Founder | Kislyn Bejagan |
| Stage | Pre-seed |
| Business model | B2B software licensing — annual and monthly plans, billed on usage |
| Sold to | Beauty brands, retailers and marketplaces. No consumer product. |
| Registered office | 352 Yishun Ring Road, Singapore 760352 |
| Laboratory | Singapore |
| Registry | ACRA, Singapore |
Founder
Kislyn Bejagan
Founder & Chief Executive
Kislyn Bejagan founded ExactHue in 2025 after concluding that the beauty industry's shade problem had been misdiagnosed for a decade. The prevailing assumption was that virtual try-on needed better rendering. The evidence pointed somewhere less glamorous: that nobody was modelling the light.
That reframing determined everything about how the company is built — a colour laboratory before a marketing team, a sampling plan before a training run, and a product that declines to answer rather than guess. It is a slower way to build, and it is the only way the result holds up when a parcel arrives.
“We are not trying to make the preview beautiful. We are trying to make it true. Beautiful is what happens when a customer opens the box and it matches.”
Focus areas
- Applied colorimetry and appearance modelling
- Representative dataset design across Fitzpatrick I–VI
- Commerce integration and returns economics
- Consent-first research practice and data governance
How we work
Four commitments we hold ourselves to.
Report the weakest band, not the average
Every accuracy figure we publish or share with a customer is broken down by Fitzpatrick band and undertone family. A single headline average would let us hide the failure mode that has damaged trust in this category for years, so we do not produce one.
Collect data with consent and pay for it
The shade library is built from sessions where contributors were informed about the purpose, compensated for their time, and free to withdraw afterwards. Withdrawal removes the sample from the library and from training. No scraping, no repurposed customer photos.
Decline rather than guess
When a capture cannot support a reliable answer, the engine returns a reason instead of a recommendation. This makes some demos less impressive. It also means a shade we do recommend is one we are prepared to stand behind.
Say what we have not solved
Heavy mixed lighting, extreme low light and certain metallic or duochrome finishes remain genuinely hard. We name these in technical calls rather than discovering them together in production.
Careers
Small team. Unusually specific problems.
ExactHue is pre-seed and deliberately small. 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 CIE technical report before writing the function.
We do not maintain a permanent listings page at this stage. If the problem described on this site is one you have thought about seriously, write to us with what you would build first and why.
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
Work with us.
Whether you are a brand carrying a returns problem or an investor looking at the category, we would rather show you the method than the pitch.