Video 03

Vexon AI Agents: From Question to Scheduled Report in Minutes

Ask a question in plain language, get a report your team can trust, then email it, schedule it or edit it just by asking. Follow a finance and operations team through seven real workflows, from suggested questions and a custom P&L to the Reporting Agent, a year-over-year expense analysis, a cash-flow paradox and slot-machine rankings.

Published 7 chapters
03 · Vexon AI agents

Demo environment: all companies, names and data shown are fictional.

What you'll see

  • Suggested questions generated from your own data model
  • A custom executive P&L, saved, pinned, exported and formatted
  • The Reporting Agent: email as PDF, weekly schedule, edit by chat
  • Proactive insights that flag what deserves a closer look
  • Deep, privacy-first analysis that explains the "why" in plain language
  • Conversational follow-ups that keep the context

Transcript

The narration, word for word, grouped by chapter. Select a timestamp to play the video from that point.

0:00Step 1: Ask Vexon (suggested questions)

This is Vexon's Data Agent. Pick a semantic model, here Casino Finance, and you're ready to ask. No blank page. Vexon suggests questions built from your own model: compare, anomaly, ranking, drill-down. One click, and Vexon picks the tables, writes the SQL, and runs it safely. Gross profit by quarter, for each casino. Ready out of the box. And every new conversation brings fresh suggestions.

0:33Step 2: Build your own report

Now, our own request: an executive P&L by month. Revenue, costs, gross profit, marginal profit and margin, for the last twelve complete months. Vexon builds the query on the semantic model, so every formula follows our business definitions. And here's our P&L, month by month. Vexon even goes further: it flags December's revenue spike, and offers to dig into what drove it.

Looks good? Save it as a report. It's now in Reports. Pin it, and it's always one click away. Open it, and export to CSV or PDF, anytime. Make it yours: format operating expenses as currency, drag columns into any order, and add a rule that highlights any expense above one million.

1:38Step 3: The Reporting Agent

Now, meet the Reporting Agent. Just ask: send me this report by email, as a PDF. Sent! Need it every week? Every Monday at nine AM, in CSV. Vexon creates the schedule for you. And to change the report, just say it. Remove direct operating costs. Vexon explains what depends on it, and removes those metrics too. Approve, and it's applied. Built, saved, shared and scheduled, in a few minutes.

2:18Step 4: Year-over-year expenses

A brand-new custom report: operating expenses by category, year over year. January to September, this year against last year, with the change in percent. Only operating accounts, depreciation excluded, sorted by this year's spend. Vexon turns it into a query on the semantic model. And there it is, as a chart. Payroll and Social Security dominates: almost half of nine hundred twenty-three million in operating spend.

Vexon is already thinking ahead: it offers to break payroll down by month, to find where the six percent increase comes from. Prefer numbers? Switch to the table. Food and Beverage grew the fastest, almost eleven percent.

3:02Step 5: The December cash paradox

Now, a real puzzle for the finance team. December was our most profitable month... so why was cash flow negative? Vexon lets you choose how deep to go: a quick interpretation, or a deep, row-level analysis. Either way, your data stays protected. Only summary statistics and a small sample are shared with the AI, and nothing is stored.

We run the deep analysis. Vexon reads every row, chunk by chunk. Paradox solved! Operations were profitable, but two one-time outflows drained the cash: a five hundred million peso financing payment, and forty-six million invested in fixed assets. Profit and cash tell different stories. Vexon connects them in seconds, in plain language.

3:53Step 6: Top & bottom slot machines

Now, Casino Operations. Pick the data source, and the suggested questions follow: slot rankings, occupancy, anomalies. Our request: rank every slot machine by net win, for the last complete quarter, with game, casino, area, hold and occupancy. Top ten, and bottom ten. This one is a heavy query, about a minute of work... so let's skip ahead.

Here's the ranking, with every detail per machine: game title, casino, hold and occupancy. And Vexon spots the star right away: machine S D Q, zero zero zero one, made four point one seven million, nearly double the next one.

4:35Step 7: The standout machine

Now, let's just chat. Which slot machines made us the most money last quarter? Vexon keeps the context of the conversation, so follow-up questions just work. The top ten by net win. Santo Domingo holds seven of them; Punta Cana, the other three. And a quick follow-up: what's the top one, and where is it located? Only one row in scope, so the interpretation is instant, and just a summary is shared with the AI.

The standout machine: S D Q, zero zero zero one, an Ocean Pearl Link in the High Limit area, at Brisalia Casino Santo Domingo. From a full ranking to one precise answer, in plain language. That's Vexon.

Your data, your questions

See Semavex on your own data

Book a demo and see how Vexon answers questions in the language of your business.

All videos
Vexon AI Agents: From Question to Scheduled Report - Semavex