AIToday
Large Language ModelsAI Business & IndustryASCII.jpPublished: Nov 22, 2026, 13:01 JST

Gomumari AI agent picks government agency targets from Nov 22

Send on LINE
Gomumari AI agent picks government agency targets from Nov 22

3 Key Points

  1. What happened

    From November 22, 2026, Gomumari's target-list AI agent will read a customer's past won-agency list — around 30 entries suffices — and rank public-sector agencies posting tenders with a per-agency selection reason.

  2. Why it matters

    The agent turns an individual bid-tracker's know-how into a shared, written selection standard, so a successor can start from the same criteria instead of rereading tender notices from the top.

  3. What to watch

    The service starts at ¥500,000 for initial setup (4 weeks or more), then from ¥100,000 per month.

WHO IT HITSSales teams at companies that sell to government agencies and municipalities — especially those with about 30 past public-sector wins — can hand a successor a written selection standard instead of relying on one person's experience.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Context & Analysis

Gomumari's new capability builds on its existing target-list AI agent, which was first offered in October 2026. The company says it tested the approach on real data: for one month (October 2026), it pulled 8,916 tender notices from 1,055 agencies across all 47 prefectures via the official public-procurement portal's search API. It also noted that of the 1,294 notices located in Tokyo, only 83 were issued by the prefecture or its wards and municipalities — the rest came from national agencies, independent administrative institutions, and national universities — illustrating why counting notices by location does not reflect local government demand. That finding underpins the design choice to select agencies based on conditions common to a customer's actual won agencies rather than on notice volume.

The agent processes data in four steps: intake, analysis, list creation, and output. It reads conditions common to won agencies, explains why those conditions matter, then selects and ranks agencies from those posting tenders, attaching a reason to each one. The output includes the list, an analysis memo, the selection criteria, and CRM-ready data. Gomumari says the list and criteria are refreshed as new wins accumulate, keeping them aligned with current business.

The company also offers the same target-list service as a one-off deliverable (from ¥200,000, delivered in 7 business days after materials are ready) for customers who prefer a productized output over running the agent in-house. Gomumari positions itself as an FDE (forward-deployed engineering) company that builds AI agents tailored to specific operations.

FAQ
What do customers need to provide?
A list of past won agencies (names alone are enough) plus product and target-area information. About 30 won entries are enough to derive conditions, and no format standardization is required.
How much does it cost?
Initial setup starts at ¥500,000 with a period of 4 weeks or more, followed by monthly operation from ¥100,000.
What does the output include?
A prioritized list with reasons, a won-analysis memo, the selection criteria themselves, and data in a format that can be imported into a CRM.
Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleMarvell, Unity top profitable AI stock screen after AI selloff