The Pulse
ThinkingAI adds AI Monitor to its Agentic Engine platform
ThinkingAI has made AI Monitor available to help administrators oversee AI use and find repeated instructions. The company also detailed security measures and a customer-hosted deployment option for Agentic Engine.

Administrators can now use AI Monitor in ThinkingAI’s Agentic Engine to review internal AI use and operate usage rules, the company said in a November 20 release. The feature also detects repeated text in instructions to AI, identifying candidates for reusable tasks or workflows.
Source prtimes.jp
ThinkingAI released the feature alongside details of security measures for Agentic Engine, its enterprise AI agent platform. The company offers both a software-as-a-service environment and a custom deployment on a customer’s own infrastructure.
Finding repeated instructions
AI Monitor has two roles: giving administrators visibility into internal AI use so they can operate usage rules, and detecting text repeatedly entered in AI instructions. The company says repeated prompts can include common assumptions or standard directions written again by different departments.
Detected repetitions can be candidates for automated tasks or Skills, which are components that package business procedures. ThinkingAI says calling those procedures when needed can reduce the work of rewriting instructions and the tokens consumed by them.
Hosted security measures
ThinkingAI says it maintains and operates the software-as-a-service environment, applying security patches to server operating systems regularly and keeping middleware access permissions from being exposed externally. High-urgency vulnerabilities are corrected promptly, according to the company.
Vulnerability scans take place when major versions are released, at least once a year, and penetration tests are conducted regularly. The company also lists malware protection, access restrictions using security groups and firewalls, and a web application firewall among its measures.
The company also says Agentic Engine supports ISO 27001 for information security management, ISO 27701 for privacy information management and the CIS Benchmark for system and cloud configuration.
Deployment options
A separate custom deployment plan places Agentic Engine on the customer’s own infrastructure. In that arrangement, data, models and inference are managed within the customer’s environment, and data does not leave it, according to ThinkingAI.
ThinkingAI offers free consultations on using AI Monitor and choosing between the hosted and custom options. The company says the appropriate choice depends on the nature of the data being handled and the organization’s internal rules.
Agentic Engine is designed to detect changes in revenue, user counts and user feedback, analyze causes and carry out measures from planning through execution. Agent features are to be provided in stages as development progresses.
Sources
- Original source · Press release Thinkingdata Co., Ltd. prtimes.jp
Related stories
- Cohere brings Crunchbase data to North for licensed customers
- Databricks puts Genie ZeroOps pipeline agent into public beta
- Google rolls out Gemini computer use on Mac for AI subscribers
- Kiro makes Conductor mode available in standard Crew builds
- Signal restricts screenshots to keep Siri and Gemini from reading chats