{"slug": "one-inference-box-dgx-spark-and-multiple-use-cases", "title": "One inference box (DGX Spark) and multiple use cases...", "summary": "A developer built an experimental RAG system, RAG-LCC, and spun out a separate GitHub repo, harinezumigel-llm-stack, to manage the multiple LLMs it relies on. The project uses separate models for prompt guarding and judging rather than a single inference model, and maintains two configuration sets for switching between use cases. The developer notes that running local coding assistance on a single DGX Spark box competes with the RAG tests for compute.", "body_md": "While developing my [RAG-LCC](https://github.com/HarinezumIgel/RAG-LCC) experimental RAG, I had multiple use cases:\n\n**RAG-LCC uses LLMs for:**\n\nAlthough some of the tasks can be handled by the same LMM it is good practice to have other models then the inference one for prompt guard and judgement.\n\nI use 2 sets of configurations, which I want to switch as convenient as possible.\n\n**Coding assistance:**\n\nI tried different models, which would be a separate discussion. But when I use local coding assistance, I cannot run the RAG tests because my Spark DGX can't provide it.\n\nBuilding a RAG system taught me that managing the LLMs I use is an important part of my development flow. What started as some lines of python code has become a GitHub repo on its own.\n\nIf you want to have a look: [harinezumigel-llm-stack](https://github.com/HarinezumIgel/harinezumigel-llm-stack)\n\nFeedback and ideas are welcome.", "url": "https://wpnews.pro/news/one-inference-box-dgx-spark-and-multiple-use-cases", "canonical_source": "https://dev.to/harinezumigel/one-inference-box-dgx-spark-and-multiple-use-cases-5h9d", "published_at": "2026-11-22 11:05:44+00:00", "updated_at": "2026-11-22 11:12:19.173991+00:00", "lang": "en", "topics": ["large-language-models", "ai-agents", "ai-tools", "ai-infrastructure", "mlops"], "entities": ["RAG-LCC", "harinezumigel-llm-stack", "DGX Spark", "GitHub", "Nvidia"], "also_reported_by": [], "takeaway": "Splitting prompt-guard and judging work onto separate models from the main inference model, and keeping two switchable configuration sets, makes a multi-model RAG stack easier to manage.", "recommendation": {"advice": "Run prompt guard and judging on separate models from your main inference model, and keep distinct switchable config sets per use case so a single box can serve multiple workflows.", "playbook": "agent-architecture", "playbook_url": "https://wpnews.pro/playbooks/agent-architecture"}, "alternates": {"html": "https://wpnews.pro/news/one-inference-box-dgx-spark-and-multiple-use-cases", "markdown": "https://wpnews.pro/news/one-inference-box-dgx-spark-and-multiple-use-cases.md", "text": "https://wpnews.pro/news/one-inference-box-dgx-spark-and-multiple-use-cases.txt", "jsonld": "https://wpnews.pro/news/one-inference-box-dgx-spark-and-multiple-use-cases.jsonld"}}