Hired announced a partnership with Fiddler to increase transparency and mitigate bias in its tech talent marketplace, which uses its proprietary AI Intelligent Job Matching technology to match candidates with relevant jobs at the world’s most innovative companies. All of Hired’s AI was built in-house, and the company is layering Fiddler’s Explainable AI Platform on top of its proprietary models to generate deeper insights into how its algorithms make decisions.
HR Technology News: TecHRseries Interview with Deborah Kearney, Global Senior Director of Human Resources at Cygilant
“At Hired, our mission is to match people with a job they love, and doing that at scale requires advanced technology like AI. Fiddler helps enhance our understanding of the AI algorithms at the heart of this candidate matching process by comparing these insights and explanations with our internally developed solutions to empower our data science and curation teams,” said Mehul Patel, CEO of Hired. “With Explainability, our team can build trust with our algorithms and further our commitment to building a truly equitable future. I am excited about Fiddler making AI decisions more explainable across the industry.”
One of the biggest promises of AI is its ability to make objective, data-driven decisions, but without visibility into how these algorithms work, businesses run the risk of using sub-par models that could actually increase bias. Fiddler’s Explainable AI Platform monitors, explains, and analyzes model performance to provide businesses with rich explanations of exactly why a model generated a particular output, and immediately flag any potential bias.
HR Technology News: 5 Tips On Creating An Efficient Virtual Onboarding Process
Fiddler’s technology integrates seamlessly with Hired’s user interfaces to provide real-time reporting on how their AI models are working. It’s specifically used for the following purposes:
- Provide real-time model performance monitoring to help data scientists and ML engineers refine models
- Enable data scientists and ML engineers to identify features contributing to a performance dip using explanations
- Generate explanations for curators, data scientists and other stakeholders on the key drivers of specific candidates’ assessment to maintain a high-quality matching process
- Provide concrete feedback with candidates to help them improve their profiles
- Generate explanations to help companies understand why specific candidates were matched with a Hired Position
HR Technology News: TecHRseries Interview with Kelly Lockwood Primus, CEO & President, at Leading Women