Alpy Evaluation
#1 v Observabilita a evaluaceCrowdworks · 2× · naposledy 18. 8. 2026
Alpy Evaluation is a module of the Agentic AI platform "Alpy" from Korean AI-Tech company Crowdworks, unveiled on February 11, 2025. The solution quantitatively and qualitatively assesses the reliability and security of AI services (LLM, RAG, AI-Agents) and is divided into a Safety Evaluation (preventing undesired/harmful responses) and a Performance Evaluation (verifying that responses occur as intended). The Safety Evaluation component combines human Red-Team assessments with model-based automated evaluation, including via the partner solution "AIM Supervisor" from AIM Intelligence. Specific pricing, deployment options, protocol compatibility, or throughput/latency values are not disclosed on the official documentation/product page.
Vlastnosti
| Deployment (Self-Hosted/Cloud) | Provided as a solution/service via Crowdworks (end-to-end enterprise AI solution), no public info on self-hosting |
| Platform | Module of the Agentic AI suite "Alpy" (alongside Knowledge Compiler, RAG, LLM Engine, Agent Studio, Admin) |
| Release Date | February 11, 2025 (launch of Alpy platform incl. Alpy Evaluation) |
| Supported Models/Providers | Model-agnostic; Alpy LLM Engine supports e.g. GPT-4o, Llama, HyperCLOVA X, DeepSeek R1 (in preparation) |