Job Requisition

Hiring needs live in managers' heads. When someone quits, the manager walks to HR and says 'I need to replace Jim.' There's no written description of what the role requires, what skills are needed, or what the position pays. The recruiter and manager negotiate the role description in real-time over coffee.

AI cannot assist with recruiting because no job requirements exist in any system. There is nothing to screen candidates against, nothing to post, and nothing to track.

Hiring managers email HR with requests: 'Need a CNC operator, 2nd shift, ASAP.' Some include job descriptions copied from the last time they hired for the role. Others are a single sentence. The recruiter creates a posting from whatever the manager provided, sometimes filling in qualification requirements from memory. No two requisitions contain the same information.

AI can parse incoming requisition emails and attempt to standardize them, but the variable format and incomplete information mean reliable candidate-to-requisition matching or pipeline tracking across positions is impossible.

A standard requisition form exists in the ATS. Every request includes required fields: job title, department, reporting manager, required qualifications, preferred qualifications, compensation range, and approval chain. Recruiters can search and sort open requisitions by department or status. But the requisition is static — once submitted, updates happen through emails and side conversations that never make it back into the system.

AI can generate pipeline reports and perform basic candidate-to-requisition matching on listed qualifications. Cannot optimize sourcing strategy or predict time-to-fill because the requisition record doesn't capture the dynamic context of the hiring process — how many candidates at which stage, where bottlenecks form.

Job requisitions are managed in an ATS with full pipeline tracking. Each requisition shows candidate counts at each stage, sourcing channel performance, average time per stage, and interview feedback summaries. Recruiters can query 'show me all engineering requisitions open more than 45 days with fewer than 3 qualified candidates in the pipeline' and get a reliable answer.

AI can identify pipeline bottlenecks, predict time-to-fill from historical patterns, recommend sourcing channels by role type, and flag at-risk requisitions. Cannot automatically adjust job descriptions or compensation ranges based on market feedback because that context lives in separate systems.

The job requisition is a schema-driven entity with formal relationships to compensation benchmark surveys, skills taxonomy nodes, workforce plan positions, and org hierarchy slots. An AI agent can ask 'which open requisitions target skills where internal bench depth is below our succession planning threshold and market compensation has increased more than 10% year-over-year?' and get a structured, integrated answer.

Autonomous recruiting optimization — AI can recommend job description adjustments, compensation range revisions, and sourcing strategy shifts based on integrated market, organizational, and pipeline signals. Routine requisition approvals and candidate shortlisting can be fully automated for standard roles.

The job requisition is a living entity that evolves in real-time. Candidate applications trigger automatic pipeline updates. Market compensation shifts trigger automatic range recommendations. Workforce planning changes trigger automatic priority re-scoring. The requisition is not a document someone filed — it is an active participant in the hiring process that keeps itself current and contextually rich.

Fully autonomous talent acquisition for standard requisitions. AI manages the entire lifecycle from requisition creation through offer generation, adjusting strategy in real-time based on continuous market and pipeline signals.

Ceiling of the CMC framework for this dimension.