Developer Hub
Documentation for teams building with Equitista
This documentation is designed for developers and analysts implementing automated diligence, research, and reporting workflows.
Overview
Equitista helps investment teams convert unstructured diligence material into verifiable, decision-ready intelligence.
- Purpose-built for analysts, deal teams, and engineering teams building workflow automations.
- Every generated insight is traceable to underlying source evidence.
- Designed for fast iteration from ingestion to report delivery.
Core Concepts
The platform is organized around a predictable set of primitives that map directly to your diligence lifecycle.
- Workspace: isolated operating boundary for users, keys, and policies.
- Data Room: uploaded source artifacts and ingestion jobs.
- Entity: company, deal, or market object being analyzed.
- Run: asynchronous agent workflow producing structured outputs.
- Report: consolidated narrative + metrics + citations package.
Security and Compliance
Security controls are designed around traceability, least privilege, and operational auditability.
- Role-based access with scoped workspace permissions.
- Source-level citation lineage for every extracted claim.
- Request IDs and webhook signatures for operational observability.
Best Practices
Follow these patterns to improve reliability, speed, and analyst confidence.
- Use idempotency keys on ingestion and run creation to avoid duplicates.
- Persist webhook delivery outcomes and retry failures with backoff.
- Keep source metadata rich (sector, period, geography) for better retrieval quality.
Troubleshooting
Operational issues usually come down to malformed inputs, auth scope mismatches, or timing assumptions.
- 401/403: confirm key scope and workspace context.
- 422: validate payload fields and file constraints.
- 429/5xx: use retry strategy with exponential backoff and request ID correlation.
Getting Started
Follow this setup sequence for a clean first implementation from API access to report retrieval.
Step 1
Create a Workspace
Set up team boundaries, invite users, and define role permissions.
Step 2
Generate an API Key
Create a scoped developer key and store it in your secrets manager.
Step 3
Upload Source Files
Ingest PDFs, spreadsheets, and transcripts to initialize your data room.
Step 4
Start an Analysis Run
Launch a run with clear objectives (e.g., risk scan, earnings summarization).
Step 5
Fetch Reports + Citations
Retrieve structured outputs and use citation references to validate findings.
Data Model and Lifecycle States
The core entities below define how ingestion, analysis, and reporting states flow across the platform.
| Entity | Purpose | Lifecycle |
|---|---|---|
| Workspace | Top-level tenant for users, keys, and policies. | active -> suspended -> archived |
| Data Room File | Raw source artifact used by runs. | uploaded -> processing -> indexed -> failed |
| Analysis Run | Agent execution with deterministic run status. | queued -> running -> completed/failed |
| Report | Human-readable output with linked evidence. | draft -> ready -> versioned |