Nearsure: An In-House Gen AI Agent that Transformed Our Hiring Process

Nearsure’s capacity team plays a pivotal role in sourcing and onboarding top-tier talent for both internal operations and client projects.

We needed a solution to identify the best candidates quickly and accurately while navigating our growing database. Traditional manual processes, which often require days to deliver viable candidates, proved both time-intensive and inadequate for addressing complex client requirements. Our objectives were clear:

• Accelerate candidate discovery: Identify the best candidates quickly and accurately while navigating our growing database.
• Enhance match quality: Analyze complex job descriptions and match them to candidates from an expansive database of 150,000 profiles, which grows by 3,000+ on average every month.
• Improve productivity: Empower recruiters to focus on high-value strategic initiatives.

We developed an AI agent integrated into our internal system. This AI agent uses natural language processing (NLP) to analyze job descriptions, extract key requirements, and query our database. Through a sophisticated ranking heuristic, it evaluates and ranks potential candidates based on role alignment.

The results include a detailed breakdown of each candidate's strengths and weaknesses, offering immediate value to our recruiters by enabling faster, more informed decision-making.

We collaborated with Microsoft’s AI Co-Innovation Labs for an intensive one-week in-person engagement, resulting in a powerful proof-of-concept solution utilizing Azure Open AI, Azure AI Search, and Azure Document Intelligence to implement intelligent search capabilities on our data modeled in Microsoft Fabric.

By leveraging Azure Open AI and other Azure services integrated with Microsoft Fabric, our system now analyzes job descriptions and matches them with ideal candidates from our OneLake curriculum database. This solution efficiently processes thousands of resumes, ranking candidates based on their fit for job criteria and providing insights into their strengths and weaknesses for the intended role.

Through an iterative development process shaped by real-world feedback from our capacity team, we created a solution that addresses our biggest challenges: time-to-hire and match quality. This innovation transformed a previously time-consuming, manual process into a streamlined, AI-driven workflow, significantly enhancing our recruitment efficiency and effectiveness.

• 30% overall hiring process time reduction

• 50% faster delivery time for first time viable candidates

• 40-hour tasks elimination