The Job Description & Skills DNA Agent automatically transforms any role definition or job description into a structured, data-driven competency profile behind the scenes.
By instantly analyzing real-time market data alongside your specific requirements, it maps core competencies, prioritizes must-have skills, and calibrates experience levels to reflect current talent realities.
Rather than treating a job description as a static text document, this connected system creates a unified source of truth that keeps hiring managers aligned upfront and automatically powers your candidate matching, scorecards, and submissions with consistent sourcing intent from start to finish.
Using the Job Description & Skills DNA Agent While Creating a Job
While creating a job, select the "Let Loxo handle this" option under Talent Discovery
Click on Customize Agents
The Job Description Agent will be toggled on by default
Select/deselect any other agents you wish to utilize on this search and click Create Job
Using the Job Description & Skills DNA Agent for an Existing Job
Once a job is created in Loxo, follow the steps below to add a job description with the Job Description & Skills DNA Agent:
Within the job, click on Manage near the top right of the page
Scroll down to the Job Description section.
To use the Job Description Agent, click on Write with AI just above the text box
The Job Description Agent may as you to provide a company URL if the profile is not already in your database or Loxo Source. The AI agent uses this link to analyze the business and refine candidate requirements for the role.
The Agent will take a short period of time compiling it's data and returning results
If you'd prefer not to have our AI write the job description, write or copy/paste your own.
Frequently Asked Questions
Q: Don’t all recruiting tools already use AI to write job descriptions?
A: Some, but they only generate text, which other systems then have to break down into messy keywords for matching. Our system converts the job description into a structured "Skills DNA" profile first, ensuring matching is based on standardized data rather than keyword guessing.
Q: What if my recruiters just paste in client JDs instead of building them from scratch?
A: That is perfectly fine and expected. The system automatically reverse-engineers pasted JDs to extract the structured Skills DNA, ensuring accurate matching even if the source text is messy or incomplete.
Q: Will the AI hallucinate or invent requirements, comp or benefits?
A: No. Hard data like compensation, compliance, and EEO language are pulled strictly from your explicit inputs, market data or pre-approved templates - never fabricated by the LLM.
Q: Is this compliant, and how do you mitigate bias?
A: Yes, the system is fully compliant because matching and scoring are deterministic and explainable, not hidden in a black box. Because evaluation logic is separated from the text and relies on structured criteria, every decision is fully auditable.
Q: How is this any different from a ChatGPT prompt I could write myself?
A: A prompt gives you a one-off, static piece of text. This system creates a permanent, versioned and editable data object that integrates directly with your workflow and powers all downstream sourcing and matching.
Q: What if the hiring manager’s written JD doesn't match what they said during intake?
A: The system translates both inputs into structured data, mapping them against each other to highlight discrepancies immediately. This allows you to catch and resolve misalignments before wasting time on sourcing.
Q: Can recruiters edit the Skills DNA if the system misses something?
A: Yes, everything is fully editable. In fact, recruiter edits are encouraged because they train the system to better understand your specific team's preferences over time. This can be done by adjusting the job description itself.
Q: How does this handle thin job orders or messy VMS data?
A: It handles them easily. Even with minimal input - like just a job title and a pay rate- the system can extrapolate a baseline structured profile so you can still source effectively where traditional keyword systems fail.