【CAPER Co., Ltd.】AI Agent Saves 80 Hours of HR Labor per Month! "OpenSeek" - A Direct Recruiting Support Platform Drastically Enhancing Hiring Power for Tier-2 Companies Struggling with Brand Awareness

No Excellent Talents Come by Just Waiting. The Fatal Challenges of Direct Recruiting: Follow-up Leakage and Lack of Resources

Nice to meet you all. My name is Shun Yamashita, Representative Director of CAPER Co., Ltd. Thank you very much for having me today.

We are a startup dedicated to solving \"human\" issues in corporate recruitment using AI agents. Following the wonderful presentation about engineer training earlier, we specialize in the preceding stage: \"recruitment.\"

Currently, many companies are facing serious resource shortages in recruitment. As a premise, excellent talents will never come to you if you just post a job opening and wait. Therefore, \"direct recruiting (scouting),\" where companies proactively approach candidates themselves, has become essential. However, no matter whether it is a startup, a small/medium-sized enterprise, or a large corporation, HR departments are severely short-handed. Furthermore, because candidates' job-seeking periods have become longer, the number of tasks and follow-up contacts (actions) has increased, making it impossible for HR to keep up.

This is based on my own past entrepreneurial experience. There was an outstanding person I wanted to hire. I found them on X (formerly Twitter), contacted them, and had an interview. However, I forgot to follow up with them due to the busy daily workload. By the time I realized it, the candidate had already decided to join another company.

This kind of \"opportunity loss due to missed follow-ups after an approach\" is a common pain point that many of you have probably experienced, not only in recruitment but also in sales. Even if you try to manage tasks using Notion or Google Calendar, it is extremely difficult for humans to meticulously follow up on every candidate. To replace this \"meticulous and polite approach activity that humans cannot fully execute,\" we developed a unique AI agent called \"OpenSeek.\"

Autonomously Listing Candidates from Over 10 Sites. A Dedicated AI Secretary Supporting HR "Between Full Automation and Full Check"

\"OpenSeek\" is a service where you simply register the candidates you want to recruit via CSV or other formats, and the AI completely manages and automates the subsequent complex communications.

Furthermore, it also has a \"talent pool expansion function\" to expand your target pool. By training the AI on \"what kind of talent your company is looking for\" in natural language, the AI automatically creates optimal search criteria. Then, using our proprietary \"Deep Research\" feature, it crosses and crawls over 10 major job sites and SNS platforms on the internet to automatically generate a list of the most suitable candidates. HR only needs to review the list and decide whether to add them to the talent pool.

After building the pool, the AI identifies optimal times when candidates are likely to reply or show high career-change intent based on data, and continues to approach them with the best message and timing.

What we focused on the most here was designing the system to be \"between full automation and full check (semi-automatic)\" rather than outsourcing all communication to AI. Under the current level of generative AI technology, fully automating all message sending carries risks like brand damage and is not realistic.

Therefore, in \"OpenSeek,\" the AI prepares message drafts and timing management for each candidate's status (e.g., scheduling interviews). However, the final sending button is pressed by a human. HR reviews the high-quality text prepared by the AI and can execute sending tasks one after another with just a click. This enables an operation as if a highly capable \"dedicated secretary\" was attached to the HR department.

By having the AI work up to 600 hours per month on behalf of humans, it reduces HR labor by approximately 80 hours per month, allowing human resources to focus on core tasks such as interviews and direct conversations.

Rapid Business Launch to ARR 30 Million Yen. Direct Hit on "Tier-2" Companies with Low Awareness but Strong Hiring Needs

\"OpenSeek\" was just launched in April 2026, but we have already signed contracts with about 20 companies, and recruitment results are already appearing at a fast pace. Currently, our ARR (Annual Recurring Revenue) has reached 30 million yen on an order basis, and we are rapidly expanding our business with the goal of achieving ARR 300 million yen by next September.

Our product is hitting particularly hard with \"Tier-2\" players in various industries. Specifically, these are upper and mid-scale SIers and mid-sized IT consulting firms that have large annual hiring volumes but suffer from low brand awareness compared to top-tier companies, which leads to low reply rates and weak direct recruiting power.

Top-tier companies with high brand awareness receive replies from candidates just by sending a scout email, but Tier-2 companies do not. That is why thorough follow-ups and optimal timing management by AI agents become a powerful weapon to increase hiring probability.

In terms of performance, compared to existing major direct recruiting platforms like BizReach, the interview transition rate remains above 10%, and the outcome probability (ease of achieving results) is about 4 times higher. We offer this powerful solution starting from a free plan to an entry plan at just 50,000 yen per month.

Becoming a Platform Managing 20 Million Talents. Outgrowing BizReach by Integrating Analog Recruitment Agencies

As a future roadmap, we have already built a system that can search and approach approximately 10 million people from open data on the web. However, our ambition does not stop there.

Our next target after open data is the integration of \"mid-sized and small-scale recruitment agencies.\" Actually, most recruitment agencies in Japan are small and their business processes are still extremely analog. We will provide a high-function CRM equipped with AI to these agencies \"completely free of charge.\"

This will automate their analog candidate management and matching tasks with companies using AI, drastically improving their productivity. At the same time, the private talent data held by these agencies will connect to our platform.

Ultimately, by combining open data and agency data, we will build a giant career platform managing 20 million talents in Japan. Backed by this data, we aim to defeat BizReach and Recruit, and grow into a platformer dominating the Japanese recruiting and career market.

Currently, we are raising funds. While we are already in talks with lead investor candidates, we would also love to build relationships from this round with investors who can invest larger amounts in subsequent rounds.

Our team consists of 3 highly capable members. Engineers who previously served as tech leads at BizReach or AI heads at Wantedly lead product development, and I am driving the business as a second-time entrepreneur. Please support our challenge to drastically transform the productivity of recruitment. Thank you very much.

Q&A and Feedback

Commentator (Mr. Fukuda): Thank you, Mr. Yamashita, for the excellent presentation. The growth speed is very fast and the business launch is clear. The concept of an AI agent performing all the scouting work and preparation on behalf of HR fits perfectly with the market pain points of direct recruiting today.

I have a question: recently on X (formerly Twitter) and LinkedIn, there has been a flood of automated DMs recommending side jobs or career changes, and recipients are starting to ignore them, thinking \"this must be AI.\" How do messages sent from OpenSeek appear to candidates? Also, please tell us about your efforts to eliminate the mechanical feel (coldness) of AI.

Mr. Yamashita: Thank you for the question. That is precisely the most important point we focused on when designing this product.

To put it simply, messages are not sent from a virtual AI account, but from the \"personal account\" of the actual HR staff or executive of the company using our service. The AI functions as their \"agent\" behind the scenes, supporting message creation and timing.

Also, as shown in the demo earlier, the AI does not directly shoot messages to candidates automatically. The AI deep-researches the candidate's background and public info to draft a highly personalized message at the best timing. HR must review the draft on the screen, and the send button is pressed only when there is no problem, maintaining a process where humans intervene. This completely eliminates the unnaturalness and spam-like feel of AI, delivering a warm approach that shows the candidate \"they contacted me after looking at my profile properly.\"

Mr. Fukuda: I see. The draft is personalized by AI but sent from the personal account of actual HR after human review. In that case, candidates are less likely to think it is spam, and there is no worry of damaging the brand.

Regarding this \"human intervention\" process, how much HR time is actually saved? If HR has to spend a lot of time checking and sending, it defeats the purpose. Can you tell us about the quantitative results of productivity improvement?

Mr. Yamashita: Quantitatively, companies actively recruiting have achieved a reduction of \"about 80 hours per month\" per HR staff.

The reason for such a dramatic reduction is that the AI automatically completes the prep work (scheduling and drafting messages) according to each candidate's status, which is where HR usually spends the most time.

When HR opens the screen, the candidates to approach today and the personalized messages drafted for each are already lined up. HR just reviews them from the top, makes minor adjustments, and sends them with a simple click. It behaves like a dedicated AI secretary, reducing overall scouting labor by about 80% and allowing HR to spend their time on core tasks like interviews.

Mr. Fukuda: A reduction of 80 hours per month has a huge impact. For Tier-2 companies with limited resources, this efficiency and timing management is a powerful solution.

Regarding this funding round, aiming for a platform of 20 million talents by integrating recruitment agencies with free CRM is a very interesting strategy to compete with giants like BizReach. I wish you the best of luck.

Mr. Yamashita: Thank you very much for the encouraging feedback. We will do our best to transform the Japanese recruitment market. Thank you.