Agmis Launches an AI-Powered CMT Platform to Automate Marketplace Content Moderation

Leading Baltic technology company Agmis has announced the launch of EasyCMT

The new AI-driven content moderation platform designed to tackle one of the most pressing challenges in digital commerce: the flood of unwanted or harmful user-generated content. Built specifically for marketplaces, ecommerce platforms, and classified ad sites, EasyCMT offers automated moderation of images, videos, and text to ensure consistent, policy-compliant user experiences at scale.

EasyCMT was developed by EasyFlow, the Agmis product spinoff known for its work in retail-focused computer vision solutions. Leveraging the same expertise used in AI-based in-store analytics and self-checkout fraud detection, EasyFlow now brings its image recognition know-how into the digital retail and content management space.

Real-Time Moderation for a Cleaner Marketplace Experience

The EasyCMT platform uses advanced computer vision and NLP models to scan every user submission including images, video clips, and written listings before it goes live on the platform. The AI detects and blocks content containing:

• Nudity or explicit imagery;
• Weapons or drug-related content;
• Hate symbols or extremist imagery;
• ALL CAPS posts, spammy characters, or excessive text overlays.

Text-based submissions are also screened for hate speech, offensive language, scams, and policy violations, helping to prevent fraud and protect community guidelines.

To further enhance content quality, EasyCMT includes background removal and image enhancement tools, allowing for cleaner and more professional-looking product images even when uploaded by non-professionals.

“Online marketplaces are like digital town squares,” says Simas Jokubauskas, Head of Product at EasyFlow. “They attract massive user activity and if left unmoderated, quickly descend into spam, scams, and poor user experiences. That’s why content moderation isn’t just a safety measure. It’s a growth enabler.”

Simas Jokubauskas, Head of Product at EasyFlow.

Flexible Moderation Workflows for Different Risk Profiles

Recognizing that moderation needs vary by platform, EasyCMT offers three distinct operating modes:

• Automated Mode: AI filters all high-risk content before it goes live. Moderators can review flagged content later, but the platform enforces blocks in real time;
• Assisted Mode: AI blocks only high-confidence violations. Content with ambiguous results is flagged for human moderation;
• Manual Mode: Content is allowed online by default. The AI assists by prioritizing review queues based on the likelihood of a policy breach.

Each piece of content receives a policy breach probability score, ensuring that human moderators spend their time where it matters most, not manually reviewing every single post.

“Human moderators are still essential, but AI should do the heavy lifting,” Jokubauskas adds. “Our platform reduces burnout, improves accuracy, and lets businesses scale without losing control.”

From Physical to Digital: Building on Real-World AI Expertise

EasyFlow has already delivered multiple successful AI tools in the offline world:

• ScanWatch reduces self-checkout fraud by auto-recognizing produce without manual input;
• Shelf Analytics provides real-time stock tracking and planogram compliance;
• Queue Management Systems optimize cashier shifts and reduce in-store wait times.

With EasyCMT, EasyFlow now extends this proven technology stack into the digital content space, targeting online marketplaces, community platforms, and e-commerce giants looking to protect their platforms and improve seller accountability.