# Genesis AI's $500M Runs on a Glove. Only 2 US Haptics Titles Turn Up

_Genesis AI's $500M round runs on a tactile glove. Exact haptics titles find 2 US profiles. Here is where hand-hardware engineers work and how to find them._

- Canonical URL: https://www.refolk.ai/blog/genesis-ai-500m-source-haptic-glove-tactile-engineers
- Published: 2026-11-22
- Reading time: 9 min
- Keywords: hire robotics engineers dexterous manipulation, tactile sensing engineer recruiting, haptic glove engineers, Genesis AI hiring, how to source robotics talent

If you're trying to hire the people who build Genesis AI's hand hardware, the obvious searches ("humanoid engineer," "robotics engineer") send you to the wrong people. In Refolk's index of US professional profiles, only 2 people hold an exact haptics or tactile-sensing engineer title, and both work at Apple.

That matters because Genesis AI just raised a $500M Series A at a $3.5B post-money valuation, and what investors are really paying for is a glove.

## What is Genesis AI actually paying engineers to build?

Genesis AI builds dexterous robot hands trained on data from a sensor glove that human workers wear, and its co-founder says the main bottleneck is hardware. Khosla Ventures, Premji Invest and Eclipse led the round. The money goes to robot development, production, data collection and hiring.

These details decide who you should be sourcing:

- **The glove maps 1:1:1.** Genesis describes a direct mapping between the glove, the human hand and the robot hand, so the sensor layout on the glove has to match the robot's hand.
- **The data comes from a fleet.** More than 100 workers around the world wear the gloves and headset cameras while they do their jobs.
- **The hand is complex.** Eno, the robot, has two arms with 20 actuated degrees of freedom per hand. Degrees of freedom (DoF) are the independent motions a mechanism can make.
- **Production is still small.** Genesis has built 30 to 40 Eno units so far.
- **The hands wear out fast.** According to TechCrunch, the dexterous hands last only a few weeks of regular work before they need replacing.

TechCrunch reports that co-founder Xian sees the key bottleneck as hardware capability, both in mechanical design and in the supply chains needed to produce industrial-grade machines. He and president Théophile Gervet met as graduate students in a Carnegie Mellon lab and started the company in 2024. Gervet was previously a researcher at Mistral AI. Sources disagree on the US base: some report San Carlos, at least one lists Palo Alto.

## Who else is hiring robotics hardware engineers with fresh money?

Agile Robots, Bedrock Robotics and Path Robotics have also announced rounds of $300M or more, so Genesis is recruiting against companies with as much money or more. The talent pool didn't grow when the funding did.

| Company | Round | Amount | Post-money valuation |
|---|---|---|---|
| Genesis AI | Series A | $500M | $3.5B |
| Agile Robots | Series F | $500M | $8.5B |
| Bedrock Robotics | Series C | $600M | $4.1B |
| Path Robotics | Series E | $300M | $2.45B |

On timing: Agile Robots announced its round on November 12, 2026, with SoftBank Group leading. SV Daily dates Bedrock's round to Nov 17, and Robot Magazine says October.

> **$9.9B** Raised by robotics startups worldwide in Q3 2026
>
> Up from $4.3B a year earlier, per Robot Magazine.

These competitors aren't evenly matched on recruiting. Agile Robots reports more than 3,500 staff from almost 60 nations, so it already runs a large hiring operation across many countries. A Series A company like Genesis can't outspend that on recruiters, so it has to find people the bigger rivals overlook. To see how much engineering talent costs right now, check the [engineering salary medians in the United States](https://www.refolk.ai/jobs/salaries/engineering/in/united-states) before you start pitching anyone.

> **Robotics engineering roles open now** - [see the open roles](https://www.refolk.ai/jobs/search?q=robotics&family=engineering)
>
> Live postings with robotics in the title, from the Refolk job board.

## Why does a title search find almost no haptics engineers?

People who build haptic and tactile hardware rarely put those words in their job title, so an exact-title search returns almost nobody. In Refolk's index, US profiles titled "Haptics Engineer," "Haptic Engineer," "Tactile Sensing Engineer" or "Haptics Hardware Engineer" total 2 people. Both are at Apple, one in Cupertino and one in Sunnyvale.

> **2** US profiles with an exact haptics or tactile-sensing engineer title
>
> Both work at Apple. The headline keyword "haptics" matches 4,335.

The work isn't rare. The label is. Haptics is the engineering of touch: sensing it, reproducing it, or both. Inside AR/VR and wearables orgs, the people doing it usually go by "hardware engineer," "sensor engineer," "mechanical engineer" or "input engineer." Their employer's org chart decides their title, not their specialty.

Here is the dataset, all from Refolk's index (US, November 2026):

| Search method | Matching profiles | Top employers in sample |
|---|---|---|
| Exact haptics/tactile engineer titles | 2 | Apple (2 of 2) |
| Headline keyword "haptics" | 4,335 | Daptic, Haptiq, Haptic Zones, J&J |
| Headline keyword "humanoid" | 4,638 | Figure, Persona AI |
| Keyword vs exact title (haptics) | ~2,170x | n/a |
| "Humanoid" vs "haptics" keyword pool | 1.07x | n/a |

Treat the keyword counts as upper bounds. Keyword matching is fuzzy: a 25-profile sample of the "haptics" results included a firefighter, and the "humanoid" sample included a bakery and a homebuilder. The ~2,170x gap tells you where the signal lives. It doesn't tell you there are 4,335 hireable haptics engineers.

## Why is "humanoid" the wrong filter for dexterous-manipulation hires?

"Humanoid" mostly finds people who already work at well-funded humanoid companies, which is the pool every competitor is already working. The headline keyword returns 4,638 US profiles, only 1.07x the "haptics" pool, and the employers in the sample include Figure and Persona AI.

There are two reasons this backfires:

1. **You're competing for the same people.** Every recruiter racing Genesis types the same word, and the people who answer to it already work at well-funded rivals.
2. **The word follows the trend, not the skill.** "Humanoid" describes a whole robot. Genesis's bottleneck is one part of it, the hand and its skin. A controls engineer on a walking robot may never have touched a fingertip sensor.

The [recruiters who split OpenAI's robotics team](https://www.refolk.ai/blog/openai-robotics-split-recruit-hardware-controls-engineers) ran into the same thing: the hardware half of a robotics team is the half that's hardest to find by title or publication record.

## What are keyword searches actually good for?

Keyword searches are good for finding companies, not engineers. In both the "haptics" and "humanoid" samples, almost every top title was CEO, CTO, founder or SVP.

The reason is simple. Executives put the category word in their headline because it describes the company's pitch. Individual contributors describe the work itself: "designed force-sensing array for wearable input," "led tendon routing for prosthetic hand." So the keyword hits work as a list of companies. The "haptics" sample points to Daptic, Haptiq, Haptic Zones and Johnson & Johnson. Treat each of those as a place to search for engineers, not as a shortlist of candidates.

The two-step version looks like this: search the keyword to map employers, then search inside each employer for the engineers who built the hardware. Done by hand, that's a lot of tabs. [Refolk](https://www.refolk.ai/) handles it as one request: describe the hardware the person built in plain English and you get back a ranked list of named people, whatever their title says.

> **Example search:** "US hardware engineers who have built haptic gloves, tactile fingertip sensors, or prosthetic hands, excluding anyone currently at a humanoid robotics company" - [run it on Refolk](https://www.refolk.ai/s/rjtkr43jmg)
>
> A ranked shortlist of individual contributors matched on the hardware they built, with current employer and evidence links, not founders who use the word in a headline.

## Where do haptic glove and tactile sensing engineers actually work?

They cluster in five places: consumer hardware teams, tactile-sensor research programs, glove makers, prosthetics and medical devices, and a few university labs. Each has products you can name and search by.

- **Apple.** Both exact-title haptics engineers in Refolk's index work there. Expect many more under generic hardware titles.
- **Meta, GelSight and Wonik Robotics.** Meta partnered with GelSight and South Korea's Wonik Robotics to commercialize tactile sensors for AI (Digit 360 and the Allegro Hand). Alumni of that fingertip program are the closest match to Genesis's problem. Running a [proven-team diaspora teardown](https://www.refolk.ai/guides/proven-team-diaspora-teardown) on that program will turn up more of them than any title filter.
- **HaptX, SenseGlove and MANUS.** These companies make glove hardware already used to collect robot data. Sanctuary AI picked HaptX as the teleoperation partner for its Phoenix robot because haptic feedback produced better demonstrations.
- **Prosthetics and medical devices.** Meta's own researchers list medicine and prosthetics among the uses of tactile sensing. Prosthetic-hand engineers already work on hand-scale hardware that has to survive daily wear.
- **Carnegie Mellon.** This is where Genesis's co-founders met. Expect Genesis to recruit hard from its own network there, so treat it as contested ground.

## Why does durability matter more than a robotics PhD here?

Genesis's hands last weeks and it has built only 30 to 40 units, so the scarce skill is shipping wearable hardware that holds up, not building demo robots. Engineers who have taken a product through manufacturing fit better than researchers who have published on grasping.

In practice, that means looking for:

- **Wear-resistant tactile skins.** E-skin is a flexible sensor layer that gives a surface a sense of touch. It has to survive abrasion, flexing and contamination.
- **Cable and tendon reliability.** At 20 DoF per hand, every flex point is a place where something can fail.
- **Manufacturing at volume.** Going from 40 units to an industrial fleet is a supply-chain problem, which is exactly the bottleneck Xian named.

The glove points somewhere else on the résumé. Genesis's glove doesn't need a robot in the loop, so building it is different from building teleoperation rigs. It is human-worn instrumentation that has to work across a fleet of 100+ workers around the world. People who have done that come from wearables, XR input devices and medical devices, not from teleoperation software.

The difficulty is that these engineers are almost all passive. They're mid-career at Apple, Meta or a med-device company, their titles don't say haptics, and they won't respond to a recruiter pitching "humanoids." A search built on job titles turns up founders, firefighters and nobody you can make an offer to, while three well-funded rivals message the same short list.

## How do you source robotics talent by what they built?

Search by the artifact, then check that the person actually built it. This works for haptics and for any hardware specialty where titles are generic.

1. **Map employers with keywords.** Use "haptics," "tactile" and "e-skin" headline searches only to list companies. Remove any company already competing with you for the same hires.
2. **Search inside each employer by artifact terms.** Use concrete words: glove, taxel, piezoresistive array, force feedback, fingertip sensor, prosthetic hand. A taxel is a tactile pixel, one sensing element in a touch-sensor grid. Piezoresistive means a material whose electrical resistance changes under pressure. People who use these words have done the work.
3. **Grade the evidence.** Treat a named product, a patent, a conference paper or a hardware repo as proof. Treat a buzzword as a claim. The [skill-evidence reference](https://www.refolk.ai/guides/skill-evidence-proficiency-tier-reference) sets out the claimed, practiced and proven tiers.
4. **Lead with the durability problem.** A wearables engineer won't be moved by "join a humanoid startup." "The hands last weeks, and you've shipped hardware that holds up" will get a reply.

## FAQ

### What kind of engineers is Genesis AI hiring?

Genesis AI's round funds hiring alongside robot development, production and data collection, and its co-founder names hardware capability as the bottleneck, in both mechanical design and the supply chain. Its hands last only weeks of regular work, so the most valuable hires are hardware engineers who have shipped durable tactile sensors, hand mechanisms and wearable instrumentation at volume.

### What does a tactile sensing engineer do?

A tactile sensing engineer designs the hardware that lets a surface detect touch, pressure and contact: sensor arrays, flexible e-skins, readout electronics and the calibration that ties them together. In practice, most of these engineers are titled "hardware engineer" or "sensor engineer" in AR/VR, wearables and medical device teams, which is why only 2 US profiles in Refolk's index carry the exact title.

### Which companies are the best sources for haptic glove engineers?

Start with Apple, alumni of Meta's tactile program with GelSight and Wonik Robotics, glove makers HaptX, SenseGlove and MANUS, and prosthetics and medical device companies. Carnegie Mellon is also a source, but Genesis's co-founders met in a lab there, so expect competition.

---

_From the Refolk blog. More posts: https://www.refolk.ai/blog/index.md_
