Productivity triples with a right-sized AI recruiter for a global RPO firm

The client

The client is a global recruitment process outsourcing firm, hired by businesses to assist them for some or all of their hiring. Their recruiters face the same pressure that every RPO firm across the globe faces: stiff SLAs set by clients, high expectations on quality and speed, and constant pressure to cover more volume at similar costs than the year before.

What was breaking

At a given instance, each recruiter carried more than 40 open requisitions, with sourcing and screening consuming up to 60% of that capacity. The firm already had legacy RPA tools and a CRM-based automation tech stack. Both added friction instead of removing it: they automated emails and tasks, while the conversations and judgment calls that move a candidate toward an offer stayed exactly where they were. Leadership saw the potential in generative AI but did not possess an in-house team with the depth to build a system designed around recruiting. What the team needed was a right-sized system that moved at its speed without losing a recruiter's judgment, working inside the ATS, email, SMS and scheduling tools it already ran on.

Why thinkbridge

Extending their existing RPA and CRM automation meant more point tools that handled the repetitive parts of the job alone, without any help left for every judgment call and every candidate conversation where it had always been.  

thinkbridge had the experts to build what the client needed, with a quick turnaround. We proposed a generative AI recruiter built upon the firm's own recruiting conversations, job requisitions and language. The recruiter was set alongside the tools recruiters already used rather than bolted on top of them.

What thinkbridge built

thinkbridge partnered with the firm's operations, talent acquisition and technology teams to build the AI Recruiter that powers their operation today. The build used the thinkOne methodology.  

Discovery came first, and the thinkbridge team and the firm’s teams worked together to understand workflows, map their process and build an as-is understanding that would act as the foundation for the build.  

Then, with a team of thinkbridge experts, the AI Recruiter was built upon the firm's own recruiting language and operating model. The tough part of the build was earning back the trust the earlier automation had enforced. Recruiters needed to see, from week one, that the agent would return judgment calls to them rather than take them over. It was then embedded into the ATS, email, SMS, WhatsApp and scheduling tools recruiters already used.  

The resulting AI recruiter handles outreach, follow-up, screening and scheduling, personalized by role, geography, experience and prior interaction. Recruiters keep the final say: every message it sends can be monitored, overridden or retrained in real time, and a weekly feedback loop built from recruiter ratings, interview outcomes and client input keeps tuning its judgment after launch. Every interaction is logged and auditable for GDPR and EEOC requirements.

The firm owns the recruiter fully and saw outcomes from week one. Over 10,000 recruiter hours were freed in 90 days. The time spent on repetitive tasks moved to higher-value work, and candidate experience held steady throughout, by the firm's own account.

What changed

3x recruiter productivity. Recruiters who managed about 40 open requisitions now manage 120 or more, with no corresponding rise in staff.

80% faster candidate response. Average time to first contact fell from 48 hours to under 10 minutes.

70% more qualified submissions. The agent's screening runs at 95% accuracy, so recruiters see more candidates who fit the role, not simply more candidates.

All customer stories