![\"Hand-drawn](\"https:\/\/rngoewtqzlssydnkvcdn.supabase.co\/storage\/v1\/object\/public\/article-images\/1bf48ea99e8c4dcab5a0988d5af28dd2.webp\"\/)
<\/p>\n\n\n\n<\/p>\n\n\n\n
You have dashboards. Good ones, probably. The trouble is that a dashboard shows you what happened, not why it happened. And when you’re an analyst or a founder who needs an answer ten minutes before a meeting, a wall of PromQL panels won’t get you there fast enough.<\/p>\n\n\n\n
This guide covers grafana ai integration<\/strong> with HotBot AI: asking your metrics questions in plain English, getting anomalies explained the way a colleague would explain them, and drafting weekly reports without hand-writing a single query. You’ll get concrete workflows, exact prompts, and a note on which HotBot model tier fits each job.<\/p>\n\n\n\n HotBot is an independent AI chat service that gives you one subscription across 800+ models from every major provider<\/a>, plus its own HotBot Chat, Chat Plus, and Chat Pro engines. It works as a Grafana ChatGPT alternative<\/a> with a wider model selection.<\/p>\n\n\n\n You don’t need to be a PromQL expert. Not needing to be one is most of the point.<\/p>\n\n\n\n Estimated Time:<\/strong> 20\u201330 minutes for setup, then seconds per question\nDifficulty Level:<\/strong> Beginner to Intermediate<\/p><\/blockquote>\n\n\n\n There are two ways to do this. Pick based on how much plumbing you want to deal with.<\/p>\n\n\n\n Option A \u2014 Copy-paste (fastest, zero setup).<\/strong> You grab a query, a result table, or a panel’s exported data from Grafana and paste it straight into a HotBot chat. This works today, needs no plugins, and respects whatever access controls you already have. Most “just tell me what this means” questions get solved this way.<\/p>\n\n\n\n Option B \u2014 Plugin bridge.<\/strong> If you want the AI to query Grafana live, there’s an open-source plugin, tamcore\/grafana-llmanalysis-app<\/a>, that connects any OpenAI-compatible LLM endpoint to your Grafana instance. It queries Prometheus, Loki, and Alertmanager in real time via tool calling. It requires Grafana \u2265 10.0.0 and an endpoint that implements the OpenAI For most analysts and founders, start with Option A. You’ll get value in ten minutes and can graduate to Option B when copy-paste starts feeling like a chore.<\/p>\n\n\n\n This is the one to start with. Here’s what to do.<\/p>\n\n\n\n A starter prompt that works well:<\/p>\n\n\n\n No plugin, no tokens, no YAML. You’re already talking to your metrics.<\/p>\n\n\n\n This is the everyday workflow, and it’s where ai data analysis with grafana<\/strong> stops being a buzzword and starts saving you time.<\/p>\n\n\n\n To see average latency you’d normally write Some prompts that work:<\/p>\n\n\n\n Model tier for this job:<\/strong> Use HotBot Chat<\/strong>, the fast everyday tier. These questions don’t need deep reasoning, and you want the answer quickly.<\/p>\n\n\n\n When you ask HotBot to write a query, give it context: your data source (Prometheus vs. Loki), the metric names, and what “good” looks like. A reference table from a Grafana dashboard tutorial<\/a> shows the patterns \u2014 error rate is Grafana Cloud has its own AI-powered anomaly detection and outlier detection that learns your historical patterns, per Grafana Labs<\/a>. Detection tells you something’s weird. It doesn’t always tell you why in language your CEO understands.<\/p>\n\n\n\n That’s where HotBot comes in.<\/p>\n\n\n\n Take a concrete case: someone pointed anomaly detection at their disk I\/O and it flagged sustained high write activity at 4 AM. The cause turned out to be a misconfigured backup job thrashing the array, as they described on Mustafa.net<\/a>. Detection caught it; understanding it was the plain-English step.<\/p>\n\n\n\n A prompt for that situation:<\/p>\n\n\n\n Model tier for this job:<\/strong> Reach for HotBot Chat Plus<\/strong> when you want step-by-step thinking \u2014 ranking causes, weighing evidence, showing its work. When the anomaly spans multiple metrics, logs, and traces and you need deeper reasoning, step up to HotBot Chat Pro<\/strong>, which brings a 1M-token context window and vision. That last part matters because sometimes the fastest input is a screenshot of the spike, and Chat Pro can read it directly.<\/p>\n\n\n\n Grafana Open Source lacks native reporting \u2014 Skedler<\/a> points this out, and it’s a real gap. You can schedule PDF exports with add-ons, but turning raw exports into a narrative is the hard part, and it’s the part nobody wants to do on a Friday afternoon.<\/p>\n\n\n\n So hand it off. Feed HotBot your week’s data and let it write the story.<\/p>\n\n\n\n Model tier for this job:<\/strong> HotBot Chat Plus<\/strong> is the sweet spot for structured, multi-step summarization without overkill. If you’re consolidating a month of data or multiple services into one exec brief, HotBot Chat Pro<\/strong> and its larger context window earns its keep.<\/p>\n\n\n\n When a third-party model is the better pick:<\/strong> HotBot gives you 800+ models<\/a> under one subscription. If your team standardizes on a specific model for report tone, or you want one known for long-context digestion, switch to it in the same chat. You’re not locked to a single model, which is the practical benefit of a real ChatGPT alternative.<\/p>\n\n\n\n Once copy-paste feels limiting, wire up Option B. With the grafana-llmanalysis-app plugin<\/a>, you go to Administration \u2192 Plugins \u2192 LLM Analysis<\/strong> and set the Endpoint URL, Model, API Key, and a Grafana Service Account Token (required for tool calling). After that, the model can query your infrastructure in real time instead of waiting for you to paste data.<\/p>\n\n\n\n If you’re on Grafana Cloud specifically, there’s also an official Grafana MCP server. A usecarly.com breakdown<\/a> explains that it exposes 50+ tools for querying dashboards, running datasource queries, and reading alerts. The catch: it only works inside a conversation you start. Nothing watches your metrics while you’re away, so don’t assume the AI is standing guard. It isn’t.<\/p>\n\n\n\n HotBot’s answer is generic or wrong.<\/strong> You probably didn’t give it enough context. Paste the actual query, the metric names, and a sample of real data \u2014 not just “explain my latency.”<\/p>\n\n\n\n The numbers don’t add up.<\/strong> Check that your CSV export didn’t truncate. Grafana’s Inspect \u2192 Data<\/strong> sometimes only shows the visible time range. Widen the range before exporting.<\/p>\n\n\n\n Plugin won’t connect (Option B).<\/strong> Confirm your endpoint implements the OpenAI You want automated, always-on monitoring.<\/strong> HotBot is conversational \u2014 it answers when you ask. For always-on detection, lean on Grafana’s own anomaly and outlier tooling (Grafana Labs<\/a>), then bring the findings to HotBot for explanation.<\/p>\n\n\n\n When you paste metric data into any AI chat, you’re sending that data to a model. Don’t paste secrets, customer PII, or credentials \u2014 scrub or aggregate first. For sensitive environments, the plugin route keeps you in more control of what leaves your instance.<\/p>\n\n\n\n One more thing worth repeating: HotBot explains, drafts, and reasons. It doesn’t run on a schedule watching your systems. Pair it with Grafana’s native alerting for the always-on layer.<\/p>\n\n\n\n Start with Step 2 today \u2014 one panel, one question, one paste. Once you trust the answers, build a saved prompt template for your weekly report so Friday takes five minutes instead of an hour.<\/p>\n\n\n\n When you’re ready, explore the full model lineup<\/a> and pick tiers per task. If you’re comparing plans, the free tier, $7.95\/week or $39.95\/quarter<\/a> options let you test everything before committing.<\/p>\n\n\n\n Your dashboards already know what happened. This gives you something that can tell you why.<\/p>\n\n\n\n No. You can ask HotBot to write queries for you, or paste results and ask what they mean. Give it your metric names and data source type for the best results.<\/p>\n\n\n\n No. HotBot is an independent AI chat service. Third-party names like Grafana identify the tools and models you can work with \u2014 they don’t imply any partnership or endorsement.<\/p>\n\n\n\n Use HotBot Chat Plus for ranked, step-by-step cause analysis. Step up to HotBot Chat Pro when you need deep reasoning across many signals or want to paste screenshots, thanks to its 1M-token context and vision.<\/p>\n\n\n\n No \u2014 it’s conversational and answers when you ask. For always-on detection, use Grafana’s built-in anomaly and outlier tools, then bring the alerts to HotBot for a plain-English explanation.<\/p>\n\n\n\n Start on the free tier with the copy-paste workflow. If you need more, plans are available at $7.95\/week or $39.95\/quarter.<\/p>\n\n\n\n Yes. One subscription gives you access to 800+ models plus HotBot’s own Chat, Chat Plus, and Chat Pro engines, so you can pick the right model per task instead of being locked to one.<\/p>\n\n\n\nWhat You’ll Need<\/h2>\n\n\n\n
Step 1: Decide How HotBot Talks to Grafana<\/h2>\n\n\n\n
POST \/v1\/chat\/completions<\/code> API.<\/p>\n\n\n\nStep 2: Set Up the Copy-Paste Workflow (Option A)<\/h2>\n\n\n\n
rate(http_requests_total[5m])<\/code>).<\/li>\nHere's a Grafana query and its results (CSV below).\nExplain in plain English what this metric shows, what's\nnormal, and anything that looks off. Assume I'm a founder,\nnot an SRE.\n\nQuery: rate(http_requests_total[5m])\n[paste CSV here]\n<\/code><\/code><\/pre>\n\n\n\nStep 3: Ask Plain-English Questions of Your Metrics<\/h2>\n\n\n\n
rate(http_request_duration_seconds_sum[5m]) \/ rate(http_request_duration_seconds_count[5m])<\/code>. One blogger described exactly this frustration on Mustafa.net<\/a>. With HotBot you just ask.<\/p>\n\n\n\nsum(rate(http_requests_total{status=~\"5..\"}[5m])) \/ sum(rate(http_requests_total[5m]))<\/code>. Paste examples like that and HotBot adapts them to your metric names.<\/p>\n\n\n\nStep 4: Explain Anomalies Like a Human Would<\/h2>\n\n\n\n
This anomaly fired at 4 AM: disk write throughput jumped\nfrom a ~10 MB\/s baseline to 180 MB\/s for 40 minutes, then\ndropped. Here's the surrounding metric data and the list of\nscheduled jobs. Walk me through the 3 most likely causes,\nranked, with what I'd check to confirm each.\n<\/code><\/code><\/pre>\n\n\n\nStep 5: Draft Weekly Reports Automatically<\/h2>\n\n\n\n
Here's this week's key metrics vs. last week (CSV attached):\nrequest volume, p99 latency, error rate, and two incidents.\nWrite a 200-word weekly reliability report for a non-technical\nleadership audience. Lead with the headline, note what\nimproved, flag the one thing we should watch, and end with a\nrecommendation.\n<\/code><\/code><\/pre>\n\n\n\nStep 6: Level Up to Live Querying (Optional)<\/h2>\n\n\n\n
Common Issues<\/h2>\n\n\n\n
\/v1\/chat\/completions<\/code> API and that tool calling is supported. Verify the Grafana Service Account Token is set correctly.<\/p>\n\n\n\nLimits and Privacy<\/h2>\n\n\n\n
What’s Next<\/h2>\n\n\n\n
FAQ<\/h2>\n\n\n\n
Do I need to write PromQL to use HotBot with Grafana?<\/h3>\n\n\n\n
Is HotBot affiliated with Grafana?<\/h3>\n\n\n\n
Which HotBot tier should I use for anomaly analysis?<\/h3>\n\n\n\n
Can HotBot monitor my metrics automatically?<\/h3>\n\n\n\n
What’s the cheapest way to try this?<\/h3>\n\n\n\n
Is HotBot a real ChatGPT alternative for data work?<\/h3>\n\n\n\n