Search and Match
- Can every output be reviewed or audited if needed?
- Does RChilli maintain reports, dashboards, and alerts for performance degradation?
- How are test results recorded and monitored?
- What safety and risk controls are in place to prevent harmful or incorrect outputs?
- What internal performance benchmarks and metrics does RChilli track?
- How does RChilli measure matching accuracy?
- How does RChilli measure Parsing Accuracy?
- Does RChilli provide Traceable Reasoning or Logs for Outputs?
- How does RChilli ensure model explainability and transparency?
- How often are bias and model quality tests conducted?
- Does RChilli use fairness metrics? If yes, which ones?
- What Datasets are used for Bias Evaluation?
- How does RChilli test models for bias across gender, ethnicity, location, age, or other sensitive attributes?
- Which sensitive attributes are excluded or controlled to reduce bias?
- How does RChilli approach fairness and bias prevention in its AI systems?
- How much time does it take to perform a search query on an Index per Job?
- What technology is used for identifying relevant and adjacent skills, and what is the baseline for skill extraction and matching?
- How does the Search and Match mechanism work, and what is the underlying basis of its operation?
- Is there a way to delete the indexed documents with the sub user ID?
- How to customize the Dynamic Weightage ?
- What are the Skill types that RChilli have or supports?
- Is there any limit to index the document?