For the complete machine-readable documentation index, see llms.txt.
← Back to projects

2025-05-18

Deviation Budget Analysis Agent

Automated pipeline that ingests spending and budgeting data, normalizes it into CSVs, prompts an LLM, and outputs a DOCX report highlighting deviations.

FinanceLLMAutomation

Outcome

Automatically produces DOCX deviation reports after merging multiple spending feeds.

Stack

PythonAzure FunctionsAzure OpenAIPandasdocx

Scope

Budget owners juggled multiple CSVs and dashboards, and deviation analysis took hours to reconcile across systems.

Approach

  • Pipeline ingests invoices, card feeds, and forecast data, normalizes the schema, and writes structured CSVs for LangChain consumption.
  • An Azure Functions workflow builds a prompt with the normalized tables and calls Azure OpenAI to narrate key deviations.
  • The agent packages the narrative into a DOCX, attaches CSV appendices, and uploads the bundle to SharePoint.

Impact

Users now receive a DOCX report in minutes, complete with deviation explanations and action items for finance stakeholders.

Projects

Need this kind of system in your team?

I help teams ship document agents, RAG copilots, computer vision pipelines, and operational automations without the usual prototype-to-production gap.

Start a project conversationivanncaamano@gmail.com