# Your monthly KPIs, with the big moves flagged. KPI reporter is an AI reporting agent for monthly KPIs. It reads your metrics table, compares the latest month with the previous month and the 3-month average, flags any move over 10%, and posts one Slack summary with the three things that most need attention. Every number comes from your data. ## Why do monthly KPI reports take so long? Because the numbers are the easy part. Someone has to copy them into a sheet, work out every comparison, decide which moves matter and write what they mean. It takes hours each month, so the report lands late, and a slow slide in churn or burn hides in a table nobody reads closely. Without it: - Someone copies the numbers into a sheet each month. - Comparisons are worked out differently each time. - A slow slide in churn hides in a table. - The report lands a week into the month. With KPI reporter: - Every metric compared the same way. - Moves over 10% flagged, with the comparison. - Three things to look at, with probable causes. - One summary on the first working day. ## What does an AI reporting agent report on? Nine monthly KPIs in four groups, each compared with last month and the 3-month average on every run. ### Revenue MRR and its growth, against last month and the 3-month average. For example: MRR, MRR growth. ### Customers Who joined, who left and how many trials became paying customers. For example: New customers, Churned customers, Net growth, Trial-to-paid. ### Margin and cash Gross margin, cash balance and monthly burn, so cost growth shows up next to revenue. For example: Gross margin, Cash balance, Monthly burn. ### Runway Cash divided by burn, or your own runway column if the table has one. It says which. For example: Months of runway, Change on last month. ## How does an AI reporting agent work? Every run follows the same five steps, then posts one summary. Each step uses one tool, and only the last one writes anything. 1. **Finds your metrics** (Tables list, reads): The monthly table you uploaded to Arclyx. 2. **Checks the columns** (Tables schema, reads): Which metrics exist, and whether runway has its own column. 3. **Compares each metric** (Tables query, reads): Latest month against the previous month and the 3-month average. 4. **Flags the big moves** (Tables query, reads): Anything over 10%, with the comparison that tripped it. 5. **Posts the summary** (Slack, posts): Three things to look at, each with a probable cause. When it runs: - Schedule: The first working day of each month, at 8:00 - @mention: @kpi-reporter how did September compare? - Webhook: Run when the month's metrics table is updated ## From one sentence to a KPI summary. Describe the job, let Arclyx draft the agent, test it once on your real table, then read what lands in #leadership. Every flag says which comparison tripped it. Example output in #leadership, "KPI reporter · September" (placeholder companies): - You · Churned customers: 14 churned, up from 9 in August (+56%). Flagged against last month. Probable cause: new customers rose sharply in July, and newer accounts churn first. Source: example-a.com/kpi-metrics - You · Trial-to-paid: Conversion 14%, down from 18% (−22%). Flagged against last month and the 3-month average. Probable cause: it's why new customers fell 12%. Source: example-a.com/kpi-metrics - You · Monthly burn: Burn $48,000, up 14% on the 3-month average of $42,100. Flagged against the 3-month average. Probable cause: gross margin fell 4 points, so costs grew faster than revenue. Source: example-a.com/kpi-metrics - You · Runway: 6.1 months, from cash of $292,800 ÷ burn of $48,000. Down from 7.2 months. Worked out as cash ÷ burn: the table has no runway column. Source: example-a.com/kpi-metrics 9 metrics · 3 flagged moves · no data gaps ## One report, four readers. The same monthly summary answers a different question for each team that reads it. - Founders: One message on the first of the month instead of a spreadsheet. - Finance: Every comparison worked out the same way, every month. - Leadership: The three things worth talking about this month. - Growth: Early warning when churn or conversion starts to slide. ## How do you automate KPI reporting with AI? Keep your monthly metrics in one table, then give an AI agent the comparisons to make and a threshold for what counts as a big move. Each month it works out every comparison, flags what moved, says what needs attention and posts one summary. In Arclyx that's five steps, starting with one sentence. 1. **Describe the job**: One sentence: which metrics, what counts as a big move, and where the summary goes. 2. **Upload your metrics**: One row per month, as CSV. Arclyx reads it as a table. 3. **Connect Slack**: Arclyx asks for the one channel it may post to. 4. **Watch a dry run**: Optional. See one full report on your data before anything is posted. 5. **Schedule it**: The first working day of each month, or any time that suits. The sentence that built it (Schedule, 1st working day, 8:00): > On the first working day of each month, compare last month's KPIs in our metrics table with the month before and the 3-month average, flag anything that moved more than 10%, and post a summary to #leadership. ### Or start from one of these sentences - Cash and runway only: "On the first of each month, report cash, burn and runway against last month and tell me if runway is under 9 months." - Board update numbers: "On the 3rd of each month, write the KPI section of our board update: each metric, its change, and a one-line note." - Churn watch: "When churned customers move more than 10% month on month, post the numbers and the probable cause to #customer-success." - Quarterly review: "At the end of each quarter, compare this quarter's KPIs with the last one and list the three biggest moves." ## AI reporting agent vs dashboards and spreadsheets. A spreadsheet, dashboard alerts and scheduled PDF reports each cover part of the job. Here is what each one does, and doesn't, do. | What you need | Spreadsheet by hand | Dashboard data alerts | Scheduled PDF reports | KPI reporter | | --- | --- | --- | --- | --- | | Compares with last month and the 3-month average | If someone builds it | No, one threshold per tile | Only what the report already shows | Yes, every metric | | Flags big moves | When someone notices | Yes, when a tile crosses a threshold you set | No | Yes, any move over 10% | | Says what to look at | Depends who writes it | No, it links to the tile | No, it sends the report | Top three, each with a probable cause | | What you get | A sheet or a deck | An in-app alert, or an email | A PDF by email | One Slack summary | | Setup | A monthly calendar slot | An alert rule per tile | A schedule per report | One sentence | Source: [Power BI docs: set data alerts](https://learn.microsoft.com/en-us/power-bi/create-reports/service-set-data-alerts), checked 17 October 2026. Source: [Data Studio docs: schedule automatic report delivery](https://docs.cloud.google.com/looker/docs/studio/schedule-automatic-report-delivery), checked 17 October 2026. ## Can you trust an AI reporting agent's numbers? Every number comes from a query on your table, causes are labelled probable, and gaps are reported, never filled. You can see every run, and every run has a spend ceiling. ### Every run is recorded - Every step, in order: Each run keeps a trace: which table it read, every query it ran, and what it decided. - Every comparison, stated: Each flag names the comparison that tripped it, and data gaps are listed, not filled. - What each run cost: Tokens and cost are recorded per run and per agent, next to your plan's limits. - A ceiling on every run: A run that would pass its dollar ceiling stops and tells you, rather than carrying on. ### It can - Read the metrics tables you share - Run read-only queries - Post to one Slack channel you name, or email you ### It can't - Change your tables - Fill in missing numbers - Connect to your billing or accounting software - Email anyone but you - Post anywhere else ## AI reporting agents, answered. ### What is an AI reporting agent? An agent that reads your numbers, works out what changed, and writes the report a person would otherwise put together. KPI reporter does it for monthly KPIs: it compares each metric, flags the big moves and posts one summary to Slack. ### Which KPIs does it report? As built: MRR and growth, new and churned customers, net customer growth, trial-to-paid conversion, gross margin, cash balance, monthly burn and runway. Change the sentence to report the metrics your own table has. ### How does it decide what to flag? Any move over 10%, against the previous month or the 3-month average. Each flag says which comparison tripped it, so you can tell a one-month blip from a trend. ### Does it explain why a number moved? For the top three, it gives a probable cause grounded in other movements in the same data, and labels it probable. It never states a cause as fact. ### How does it work out runway? Cash balance divided by monthly burn, unless your table has a runway column. The message says which one it used. ### What if a month is missing data? It says which values are missing and leaves those comparisons out. It never fills a gap or works a number out from missing values. ### Where does the data come from? A monthly metrics table: a CSV export from your accounting, billing or spreadsheet, uploaded to Arclyx as a table. It reads it with read-only queries and changes nothing. ### Can I get the summary by email instead of Slack? Yes. Arclyx can send the summary to your own inbox instead of Slack, or as well as it. The email only ever goes to you, the person the agent runs for: it has no way to email anyone else. ### Do I need to write code? No. Describe the job in one sentence. Arclyx builds the agent, connects your table and Slack, and shows you a dry run before you schedule it. ## Agents that work well with it - [Finance desk](https://arclyx.ai/agents/finance-agent): A checked month-end summary with decisions - [Bank reconciler](https://arclyx.ai/agents/bank-reconciliation): Bank matched to ledger, exceptions explained - [Feedback analyst](https://arclyx.ai/agents/customer-feedback-analysis): Customer feedback grouped into ranked themes ## Early access Build your KPI reporter. One sentence in, a working agent out. Join the beta and we'll set you up. Join at https://arclyx.ai/#join or write to hello@arclyx.ai. Last updated 17 October 2026.