AI Hiring Velocity.
How fast each AI role is heating or cooling. Posting volume, time-to-fill, salary drift, competitive intensity. Weekly updates with confidence intervals on every metric.
What it is, and why it exists.
The AI labor market moves faster than any market we track. LLM Systems Engineer roles surged from near-zero in early 2024 to a leading hire in 2026. Prompt Engineer postings peaked in mid-2024 and have been declining ever since. RAG Platform Engineer is in the middle of an inflection. These dynamics matter because they determine when offers will be aggressive, when time-to-fill will blow out, and when comp is moving up or down.
The Hiring Velocity Engine takes four observable metrics and turns them into a posture read for every role × city combination. Posting volume tells you whether demand is rising or falling. Time-to-fill tells you whether the market is supply-constrained or in equilibrium. Salary drift tells you whether comp is moving up or down. Competitive intensity tells you how many companies are bidding for the same talent.
We surface the four metrics as separate fields rather than collapsing them into a single heat index. A role with high posting volume but stable salary drift is heating in a different way than one with stable posting volume but rising drift. The first is a hiring boom (more roles, same comp); the second is a supply crisis (same number of roles, but each one is harder to fill so comp is bid up). Different actors need different reads.
Velocity composes with every other engine. Role Taxonomy provides the role identity. Comp by Stage provides the band. Hiring Velocity tells you how fast the band is moving. Skill Premiums tells you which skills are most leveraged in that band right now. Through the MCP analyze_market tool, the agent gets all of these in one Decision-Ready Response.
Sample velocity readings.
National rollup for May 2026. Hot = top quintile by combined metrics; Cold = bottom quintile. WoW = week-over-week, QoQ = quarter-over-quarter.
Methodology version 2026.2-velocity. Live data via GET /api/v1/intelligence/salaries/hubs/[hub] and GET /api/v1/intelligence/salaries/cities.
The four metrics.
Trailing 7-day count of distinct open requisitions, normalized against the trailing 90-day baseline. Reports as week-over-week percentage change. Positive growth means more requisitions opening than closing.
Median days from posting to either filled-or-pulled. Computed over the trailing 90 days of closed requisitions. High TTF means supply-constrained; low TTF means roles are filling quickly and competition is low.
Quarter-over-quarter change in the median posted salary range for the role. Smoothed with a 4-week rolling window to suppress posting-quality noise. Positive drift means comp is moving up: a leading indicator of mid-year raise pressure.
Count of distinct companies with at least one open requisition for the role in the city in the trailing 90 days. Above 10 means candidates have leverage; under 3 means companies have leverage. The threshold is exposed so customers apply their own thresholds.
Use cases.
When competitive intensity is high and time-to-fill is blowing out, recruiters can either accelerate hiring (offer aggressively to close before peers) or delay (wait for the heat to subside). The velocity feed tells them which moment they are in for each role they are filling.
Salary drift is a leading indicator of mid-year raise pressure. When drift is positive for a role for two consecutive quarters, comp teams know to budget for retention raises or risk losing key engineers. The engine's 4-week rolling window suppresses noise so teams act on signal not seasonality.
The strongest negotiation moment is when posting volume is up, time-to-fill is blowing out, salary drift is positive, and competitive intensity is high. The engine surfaces all four simultaneously so candidates can ask "is now the moment?" with a quantified answer rather than a vibe.
Through the MCP analyze_market tool, agents pull velocity for any role / city combination and produce a Decision-Ready Response with the posture read, the time-to-fill, the comp drift, and a suggested action.
API integration.
Two integration paths.
Full reference at /orbyt-intelligence/api-docs. Webhook subscriptions on velocity-change events are available on the Scale tier.
Pricing and access.
Velocity queries require the Pro tier and the market:read scope. The Scale tier adds webhook subscriptions on velocity-change events (e.g., role enters Hot status). Full pricing at /orbyt-intelligence/pricing.
See also.
Methodology version 2026.2-velocity. Last updated May 2026. Posting volume, time-to-fill, competitive intensity update weekly. Salary drift updates quarterly with a 4-week smoothing window.