Back to research
Four-Leaf Research · Issue 01

The AI-Era Hiring Index: 2026 Q2

A snapshot of 3,502 open roles at 16 AI-native and high-growth employers, captured on 2026-04-14. First issue of a quarterly series.

Four-Leaf TeamMethodology

Dataset released under CC BY 4.0. Attribution: “Four-Leaf AI-Era Hiring Index, 2026 Q2.”

Key findings

  • 45% of the 3,502 JDs in this index disclose a salary range directly in the description text. Discord leads at 99%.
  • Anthropic posts the highest median salary band in the index, at $290K to $365K (median of 307 disclosed roles).
  • Python is the most-mentioned skill in engineering JDs across the index, named in 33% of roles.
  • LLM or foundation-model experience is listed in 57% of data-and-ML JDs and 65% of research JDs. Once a specialty, now a default expectation at these employers.
  • The “remote” flag reported by ATS systems does not match JD text. Across the index, ATS flags mark 25% of roles as remote while a text scan of descriptions matches 44%. Readers relying on ATS metadata alone are getting a distorted picture.

How this was built

Four-Leaf operates a nightly scraper against public ATS feeds (Greenhouse, Ashby, Lever) for a curated index of 16 AI-native and high-growth employers. On 2026-04-14, the snapshot captured 3,502 active roles. Role family, seniority, salary range, and skill mentions were extracted from JD text using a published keyword and regex taxonomy. Full methodology, including false-positive guards and seniority heuristics, is at four-leaf.ai/research/methodology.

What AI-era employers are hiring for

Role mix across the index skews engineering-heavy, with sales and operations as the largest non-technical buckets. A sizeable “other” bucket reflects roles whose titles don’t map cleanly to a family (cross-functional, domain-specific, or ambiguous wording).

Role familyRoles% of index
engineering89125%
other78222%
sales gtm62418%
data ml3339.5%
operations2577.3%
marketing1614.6%
finance legal1464.2%
product1323.8%
research692.0%
support511.5%
people290.8%
design270.8%

Top skills in engineering roles

Mention rate across 891 engineering JDs. Skills are detected by case-insensitive whole-word matches; a skill is counted once per JD regardless of how many times it appears.

Python
33% (298)
AWS
22% (195)
LLM / foundation model experience
21% (189)
Go
19% (171)
Java
19% (170)
TypeScript
18% (158)
JavaScript
18% (156)
Kubernetes
16% (144)
React
15% (136)
GCP
15% (133)

Top skills in data and ML roles

Mention rate across 333 data and ML JDs.

LLM / foundation model experience
57% (188)
Python
44% (148)
SQL
23% (75)
AWS
16% (52)
PyTorch / TensorFlow
16% (52)
Spark
14% (46)
Java
12% (41)
GCP
12% (40)
Go
11% (35)
Figma
8.1% (27)

Salary transparency scorecard

A role is counted as salary-transparent when its description contains an extractable USD pay range (e.g. “$150,000 to $200,000” or “$150K–$200K”). Sanity bounds reject values below $30K and above $2M; context windows around each match filter out obvious non-salary dollar mentions (funding amounts, bonuses, equity refresh language). Medians are reported only where at least one role discloses a range; companies with zero disclosures appear as “not disclosed in JD” and almost certainly link to a separate comp page or pass ranges out-of-band.

CompanyJobs% disclosedMedian minMedian max
Discord7999%$196K$221K
Instacart13391%$176K$186K
Pinterest14282%$150K$300K
Anthropic42772%$290K$365K
Scale AI17170%$206K$261K
Figma15462%$153K$303K
DoorDash45361%$116K$170K
Vercel7860%$196K$280K
Airbnb23653%$191K$225K
Notion15551%$180K$220K
Datadog42844%$143K$197K
Ramp13012%$128K$182K
Stripe4951.8%$157K$229K
Spotify1760.0%not in JDnot in JD
Coinbase1470.0%not in JDnot in JD
Plaid980.0%not in JDnot in JD

Medians are computed across each company’s disclosed-range subset; the sample size varies by company and is included in the JSON stats file. A median band is not the same as a “typical offer”: JDs often cover a wide seniority span and US state differentials.

Remote-work reality check

Every row below uses two independent signals. The ATS flag is the boolean sent by the source ATS (Greenhouse, Ashby, or Lever); the text signal is whether the title, location, or description contains explicit remote-work language. At several employers the two signals diverge by more than 50 percentage points, suggesting the ATS flag is set by defaults rather than per-role policy. For reporting purposes this means neither signal alone is reliable, and we publish both.

CompanyJobsATS flag remoteText remoteHybrid in text
Stripe49515%19%2.4%
DoorDash45311%99%15%
Datadog42812%15%94%
Anthropic4277.3%9.4%100%
Airbnb23620%51%4.7%
Spotify1760.0%78%1.1%
Scale AI1710.0%11%8.8%
Notion15563%2.6%1.3%
Figma1540.0%100%18%
Coinbase14793%100%7.5%
Pinterest14258%62%43%
Instacart13392%93%2.3%
Ramp13098%4.6%1.5%
Plaid981.0%4.1%4.1%
Discord7922%22%3.8%
Vercel7830%97%71%

Largest ATS-vs-text divergence

  • Figma: ATS flag 0.0%, text signal 100% (gap of 100 points across 154 roles).
  • Ramp: ATS flag 98%, text signal 4.6% (gap of 93 points across 130 roles).
  • DoorDash: ATS flag 11%, text signal 99% (gap of 88 points across 453 roles).

How to cite this report

Research output, dataset, and methodology are released under CC BY 4.0. Attribution required.

Four-Leaf Team. (2026). The AI-Era Hiring Index: 2026 Q2. Four-Leaf Research. Retrieved from https://four-leaf.ai/research/ai-era-hiring-index-2026-q2

Companies in the 2026 Q2 index

Airbnb, Anthropic, Coinbase, Datadog, Discord, DoorDash, Figma, Instacart, Notion, Pinterest, Plaid, Ramp, Scale AI, Spotify, Stripe, Vercel. The index is curated for signal density (AI-native or high-growth technology employers with public ATS feeds). It is not a representative sample of the broader US labor market. Quarterly updates may add or remove companies based on ATS availability; the complete list per quarter is always published in the stats JSON.

Want the next issue?

New snapshots publish quarterly. Follow the Four-Leaf research hub or subscribe to get Issue 02 (2026 Q3) in your inbox.