“There are many applicants, but no one to hire.”
Searchlight AI is a startup born from the desperate voices of hiring managers. The company developed a B2B SaaS-based talent search solution that uses AI to identify "passive job seekers" who don't apply to job openings and then provides companies with a list of suitable candidates.
Searchlight AI touts its customized candidate recommendation feature, specifically optimized for startup hiring environments. While most recruiting solutions focus on proactive job seekers (voluntary applicants), Searchlight AI leverages undisclosed external data sources like Remember, LinkedIn, and press releases to include candidates with no intention of changing jobs in its recommendations.
Searchlight automatically extracts and provides "experience tags" and "competency insights" for talent. For example, it tags actual project-based competencies like "B2B SaaS experience," "0-to-1 leadership experience," and "LLM implementation experience," and automatically analyzes job fit based on these tags. This allows companies to review candidates based on quantifiable criteria rather than mere resumes.
Most recruiters had to repeatedly conduct manual searches, which typically took over 20 hours, including writing job descriptions (JDs), sourcing candidates, sending out messages, and persuading them. Searchlight, however, uses an automated system to instantly provide a list of 50 suitable candidates. Customers can access this service in a variety of ways, including one-time requests, monthly subscriptions, and linked recruitment.
To date, Searchlight has secured over 60 paying clients and is currently working with companies such as GS Retail, CJ Olive Young, Healing Paper (Gangnam Unni), Revit (Always), and East Aid. Searchlight boasts high repeat purchase and referral conversion rates across a wide range of clients, from startups at seed and Series A stages with limited recruiting resources to large corporations.
The company is currently transitioning from a list-based model to a SaaS offering. It is enhancing its product to provide at-a-glance access to detailed candidate profiles, competency-based match scores, and external data-driven evaluation reports. To achieve this, the company is developing an AI model utilizing LLM-based LangGraph technology. Specifically, the company is building a system that automatically collects, processes, and recommends candidate data based on a Retrieval-Augmented Generation (RAG) architecture.
CEO Yang Seung-mo stated, "Startups require distinctly different talent profiles at each stage of their growth. We aim to improve the quality of our hiring process by providing sophisticated recommendations tailored to the company's specific needs, rather than simply searching resumes." The team consists of co-founders with extensive practical experience in recruitment, product, and data.
The domestic recruiting market is estimated to be worth approximately 2 trillion won, with outbound recruiting accounting for approximately 35% of that. The startup talent recruitment market, in particular, is expected to exceed 10 billion won annually after Series B. Searchlight is planning a full lineup of SaaS solutions that will "provide candidate shortlists, convert applicants, and improve hiring rates." It is also developing a resume-based company recommendation feature and a personality-based performance management tool.
Searchlight AI's attempt to identify "hidden talent" and automate the entire recruitment process using AI represents a novel approach to addressing the chronic talent mismatch problem. Real-time data-driven talent search could mark a significant turning point in the future recruitment market.
Searchlight's strategy of using AI to address the structural limitations of the recruiting market—a market with a plethora of applicants but a dearth of talent—is expanding beyond simple solutions to market definition. If the automated pipeline, from list provision to improved hiring rates, is fully realized, Searchlight is likely to establish a unique position in the HR SaaS market.
You must be logged in to post a comment.