Announcing AutoGuard™
from Armilla

Protect Generative AI solutions with AutoGuard, the most
advanced firewall for Generative AI.

Groundbreaking Alignment Platform

Since 2020, Armilla AI has been the trusted leader for model testing and validation solutions. It provides a common platform for testing a target model's use case specific capabilities.

Generative AI changes the game. AutoAlign™ is Armilla's new platform for aligning generative AI against real-world requirements. AutoTune™ provides testing and hardening of LLMs. AutoGuard™ provides a production firewall to prevent harms and make generative AI solutions safe.

How It Works

Armilla's AutoAlign™ Platform generates its own synthetic prompts & expected completions for testing, and can optionally fine-tune the target model using automated feedback based on these concepts.

Throughout the process, Armilla's own AI aligns the target model towards user-specified expectations by understanding the requirements for the target mode, generating synthetic data to capture the bounds of the model's behavior, iteratively testing, discovering weak spots, and can either fine-tune the model and/or create its own guardrail model.

Benefits

Immediate Value
A wide range of out-of-the-box alignment controls available, with the ability to customize or create new controls.
High Performance
Combines the strength of guardrails and fine-tuning, enabling greater robustness than either alone can provide.
No Data Required
Concept-based approach requires minimal data, and minimal human-in-the-loop contact. Auto generates targeted synthetic data.
Automated Testing
Each alignment scenario can be tested and assessed independently. Detect bias, security holes, tonality issues, hallucinations easily.
Supports Major Base Models
Including OpenAI LLMs like GPT, Stable Diffusion image generation models, open source models, with Bard and others coming soon.