AI integration steps
The ALM functionality includes built‑in AI features that enhance team productivity when working with project artifacts. To use them, install the additional components using the following command:
cd /home/devprom/docker
docker-compose -f aitools.yml up -d
| Component | Description |
|---|---|
| mcp |
An MCP service that provides AI agents with a clear, extended API description for working with project artifacts (reading, creating, modifying), semantic search (RAG), and so on. To use it in an AI agent, connect the MCP service:
|
| chromadb | A vector database management system (vector DBMS) for storing embeddings (vectorized representations of user data). Vectors are generated and cached when data is created or modified, then used by AI functions without consuming computational resources of language models. |
| ollama |
An open AI model management service that allows connecting paid and free language models, as well as other specialized AI models. This service is designed for local use of AI functions (without internet access), but requires dedicated computing resources (e.g., GPU) to ensure acceptable model performance and quality. If external LLMs are used, this component is not required. |
| ollama_models | A temporary container that installs a free model for generating vector representations (embeddings) for evaluation purposes. An internet connection is required to download the model into the Ollama service. |
You can specify AI-model usage settings in the Administration → Settings → Application section.
