Welcoming November with a major update to TozheBot!
4 viewsExternal agents, improved data collection, community moderation, security enhancements, and much more—all included in this update!
We’re welcoming November with a major update!
We continue to update the TozheBot agent-creation platform, and we’re kicking off the new month with some major changes. Below, we’ve outlined the most interesting new features and, as always, we look forward to your feedback.
New tab - users
View aggregated data and conversations associated with a specific user. You can now track all sessions for an individual in the dedicated section. There, you can also review recent activity, access specific conversations, and see a list of the knowledge bases and tools that have been used.
Launch research with a single click
A new experimental feature for business owners. Analyze the market using publicly available data and compare it against your own business metrics, all while incorporating context from your knowledge base. The agent is already familiar with your business and supplements that understanding with facts from public sources. You can find all of this on the analytics page:
Usage limits for connectors
Set monthly token usage limits for individual connectors - whether for Telegram, Discord, VK, or external agents:
Security updates
We have updated the security measures for working with agents, including external connections to agents and internal models:
- Updated firewall rules and API-level security, including shorter key rotation intervals.
- Enhanced network-based DNS traffic filtering to improve service stability.
- Implemented Cloudflare Tunnels for added security and transitioned to post-quantum certificates.
- Updated the security and personal data handling model for OGL-Mini.
Other updates
- Agents now have an experimental tool for searching the internet.
- For Telegram and Discord connectors, community moderation features have been added (moderation rules can be configured in the connector settings), including the option to alert a human operator if rules are violated.
- Interfaces in several sections of the control panel have been updated.
- The API Reference for various platform sections has been updated.
- We have received an "AI Verified" certificate (bronze level): https://aiverified.io/business/ru/neurosell-llc-8b50df43ef32
And, of course, a new roadmap for November has already been published: