Troubleshooting Agent and Operations Agent
2:07

Troubleshooting Agent and Operations Agent

Monte Carlo 28.01.2026 29 просмотров

Machine-readable: Markdown · JSON API · Site index

Поделиться Telegram VK Бот
Транскрипт Скачать .md
Анализ с AI
Описание видео
2 minute demo covering two of the AI agents in the Monte Carlo platform to help you resolve issues quickly.

Оглавление (1 сегментов)

Segment 1 (00:00 - 02:00)

Let's take a look at the troubleshooting agent. One of three agents in production in the Monte Carlo platform. Here we can see that we have 10,000 more rows than we expected in our subscription detail table. We can see that anomalous jump here. If we go into our troubleshooting agent, immediately we have a summary of what is uh potentially the root cause of this issue. This would take on average about 19 hours for an organization to do manually. However, Monte Carlo has provided that in less than 30 seconds. We can see here that um we've the agent has gone through reasoning at all the upstream tables to examine what changes have occurred across the data the systems in the code and has found that this is potentially a code issue. Right? We have a um essentially a multiplicative join that has really created an explosion of data here. Right? And so the power here is not only have we reduced a 19-hour process into 30 seconds, but we can uh verify this. All this data uh is uh provided along with the ability to go in the Monol platform and verify it for ourselves. So we know we don't have a hallucination of any sort here. And we can also engage further with the agent and ask uh what are the next steps. This could be particularly powerful if you have a less technical user that is trying to extend in the process and see if they can either submit a ticket to the data engineering team that has a little bit more information, uh actionable next steps or maybe they even want to uh take the next steps themselves through some of the more uh simple issues. Here we can see that the uh immediate next action is to review the join logic and so we can actually ask the troubleshooting agent or the operations agent in this case um to go ahead and do that for us. So while we are looking at this join logic we've asked it to create a query to fix that issue and to keep one row per unique key combination and this could be part of a dramatically accelerated uh remediation process. The troubleshooting agent has automatically provided uh this query structure that and instructions to implement dramatically accelerating the remediation process uh even for less technical users. Very powerful value.

Другие видео автора — Monte Carlo

Ctrl+V

Экстракт Знаний в Telegram

Экстракты и дистилляты из лучших YouTube-каналов — сразу после публикации.

Подписаться

Дайджест Экстрактов

Лучшие методички за неделю — каждый понедельник