- Introdução
- Melhores práticas
- Capabilities
- Habilidades
- CLI
- Exemplos
- Visão geral de exemplos
- Receita: criar um agente
- Receita: automatizar com RPA
- Receita: insira dados em um aplicativo da Web
- Receita: extrair dados de um aplicativo de desktop
- Receita: refatorar e testar um fluxo de trabalho
- Receita: um processo baseado em fila com REFramework
- Receita: criar um Maestro Flow
- Recipe: build a Maestro Case
- Receita: app codificado e fluxo de trabalho da API
- Recipe: build a coded function
- Receita: extrair dados de documentos com IXP
- Avançado
- Ajuda
Process Mining process app authoring from a CSV or event log, covering template selection, data mapping, ingestion, transformation, and querying for throughput, variants, and bottlenecks.
A coding agent can take a raw CSV or event log and turn it into a queryable Process Mining app — mapping the data, loading it, running the transformation layer, and querying the result for throughput, variants, rework, and bottlenecks.
O que você cria
A process app moves through a lifecycle: it's created from a template and a data mapping, loaded with your source data, transformed through a dbt-based layer into the process model, and published so dashboards can query the full dataset. The agent drives every stage of this from the terminal.
- Event logs — a single denormalized log of case, activity, and timestamp uses the generic event-log template.
- Source-system extracts — a full multi-table extract from a system such as SAP, Oracle, NetSuite, ServiceNow, or Salesforce uses the matching source-system template (for example, purchase-to-pay on SAP).
Once published, the agent can query the app directly — grouped metrics, raw detail rows, percentiles, and root-cause analysis — or help you extend the data model with a custom analytical table.
A habilidade por trás disso
This work is powered by the uipath-process-mining skill, which builds a process app from your data, authors the transformation layer, and queries the result. You can see it in the Skills catalog.
Um prompt de inicialização típico
Uma primeira solicitação pode ser semelhante a isto:
"Create a process app from this event log and identify the main bottlenecks."
The agent picks the matching template, maps your columns, loads and transforms the data on a small subset first, then publishes once the model looks right.
Mais prompts para tentar
- Ingest this CSV as a purchase-to-pay process app from our SAP export and show me the average steps-to-resolution by vendor.
- Add a calculated field for total cycle time and group cases by whether they breached SLA.
- My dashboards won't open — check whether a field's data kind is mismatched.
Descubra mais
For the full picture of Process Mining apps and dashboards, see the Process Mining documentation.