How siddheshg01/argos works
This is a financial operations intelligence platform: it turns sales data into validated financial reports, root-cause findings, and forecasts, then offers analyst, policy-retrieval, multi-agent, and approval-gated workflow capabilities through a dashboard and API. The documented analytical workflow is represented alongside those product branches. Source excerpts show orchestration and selected integrations; links to unsampled subsystems are included only where the README or SOURCE INDEX supports them.
Main entry points: pipeline.py Financial pipeline · forecasting.py Financial forecasting · main.py CLI entry · rag_pipeline.py Policy RAG pipeline
Components 31
Financial analysis
- CLI entry
main.py - Dataset loader
src/financial_intelligence/loader.py - Schema discovery
src/financial_intelligence/schema.py - Data validation
src/financial_intelligence/validator.py - Data cleaning
src/financial_intelligence/cleaner.py - Feature engineering
src/financial_intelligence/features.py - Financial pipeline
src/financial_intelligence/pipeline.py - Financial reports
src/financial_intelligence/report.py
Investigations and forecasts
- Root-cause analysis
src/financial_intelligence/investigation.py - Financial forecasting
src/financial_intelligence/forecasting.py - Forecast models
src/financial_intelligence/forecast_models.py - Forecast reports
src/financial_intelligence/forecast_report.py
Analyst and agents
- LLM analyst
src/financial_intelligence/llm_analyst.py - Prompt construction
src/financial_intelligence/prompt_builder.py - Agent orchestrator
src/financial_intelligence/agents/orchestrator.py - Specialist agents
src/financial_intelligence/agents - Agent messages
src/financial_intelligence/agents/models.py
Policy retrieval
- Policy RAG pipeline
src/financial_intelligence/rag/rag_pipeline.py - Policy documents
src/financial_intelligence/rag/document_loader.py - Document chunking
src/financial_intelligence/rag/chunker.py - Gemini embeddings
src/financial_intelligence/rag/embeddings.py - Policy vector store
src/financial_intelligence/rag/vector_store.py
API and operations
- Workflow engine
src/financial_intelligence/workflow.py - Approval controls
src/financial_intelligence/approval.py - Action execution
src/financial_intelligence/actions.py - Workflow state
src/financial_intelligence/workflow_store.py - FastAPI adapter
src/financial_intelligence/api.py - Operations dashboard
web/src/App.jsx
Other components
- Finance user
- Sales dataset
- Gemini service
Connections 38
- Finance user → CLI entry: runs analysis
README.md - Sales dataset → Dataset loader: loads rows
README.md - CLI entry → Financial pipeline: runs pipeline
main.py - Financial pipeline → Dataset loader: loads data
src/financial_intelligence/pipeline.py - Financial pipeline → Schema discovery: discovers columns
src/financial_intelligence/pipeline.py - Financial pipeline → Data validation: validates data
src/financial_intelligence/pipeline.py - Financial pipeline → Data cleaning: cleans rows
src/financial_intelligence/pipeline.py - Financial pipeline → Feature engineering: engineers features
src/financial_intelligence/pipeline.py - Financial pipeline → Financial reports: writes reports
src/financial_intelligence/pipeline.py - CLI entry → Root-cause analysis: runs investigation
main.py - CLI entry → Financial forecasting: runs forecasts
main.py - Financial forecasting → Dataset loader: loads dataset
src/financial_intelligence/forecasting.py - Financial forecasting → Schema discovery: discovers schema
src/financial_intelligence/forecasting.py - Financial forecasting → Data cleaning: cleans data
src/financial_intelligence/forecasting.py - Financial forecasting → Feature engineering: engineers features
src/financial_intelligence/forecasting.py - Financial forecasting → Forecast models: evaluates candidates
src/financial_intelligence/forecasting.py - Financial forecasting → Forecast reports: writes outputs
src/financial_intelligence/forecasting.py - CLI entry → LLM analyst: runs analysis
main.py - LLM analyst → Prompt construction: builds prompts
src/financial_intelligence/llm_analyst.py - LLM analyst → Gemini service: requests generation
src/financial_intelligence/llm_analyst.py - CLI entry → Policy RAG pipeline: runs policy query
main.py - Policy RAG pipeline → Policy documents: loads documents
src/financial_intelligence/rag/rag_pipeline.py - Policy RAG pipeline → Document chunking: chunks documents
src/financial_intelligence/rag/rag_pipeline.py - Policy RAG pipeline → Gemini embeddings: creates embeddings
src/financial_intelligence/rag/rag_pipeline.py - Policy RAG pipeline → Policy vector store: stores vectors
src/financial_intelligence/rag/rag_pipeline.py - Gemini embeddings → Gemini service: requests vectors
src/financial_intelligence/rag/embeddings.py - Finance user → Operations dashboard: uses dashboard
README.md - Operations dashboard → FastAPI adapter: calls API
README.md - FastAPI adapter → Agent orchestrator: routes queries
src/financial_intelligence/api.py - Agent orchestrator → Specialist agents: dispatches requests
src/financial_intelligence/agents/orchestrator.py - Agent orchestrator → Agent messages: uses messages
src/financial_intelligence/agents/orchestrator.py - FastAPI adapter → LLM analyst: routes analysis
src/financial_intelligence/api.py - FastAPI adapter → Policy RAG pipeline: routes policy queries
src/financial_intelligence/api.py - FastAPI adapter → Workflow engine: routes workflows
src/financial_intelligence/api.py - Workflow engine → Approval controls: requires approval
src/financial_intelligence/workflow.py - Workflow engine → Action execution: executes actions
src/financial_intelligence/workflow.py - Workflow engine → Workflow state: persists state
src/financial_intelligence/workflow.py - Approval controls → Workflow state: records approvals
src/financial_intelligence/approval.py