Model serving
Ollama, vLLM, local gateways, and OpenAI-compatible profiles can support deploy-anywhere runtime choices.
OSS as strength
SENTINEL is designed to use open-weight models and open-source infrastructure responsibly, with license and provenance checks before production use.
Ollama, vLLM, local gateways, and OpenAI-compatible profiles can support deploy-anywhere runtime choices.
Qdrant, embeddings, rerankers, parsers, and source manifests support grounded answers.
Citation parsers, official source connectors, and authority verification matter more than raw model prose.
ComfyUI-style workflows, OCR, multimodal models, and provenance layers expand visual work.
SBOM, scanner, secret detection, signing, and release evidence make the stack reviewable.
Credits and license labels are public-facing parts of responsible OSS adoption.