ProSights

What do they actually do

ProSights builds an API‑first service that extracts tables, charts, and text from financial documents like PDFs and images and returns structured outputs that analysts can audit and download. Output formats include JSON and XLSX (with a focus on cell‑level structure and chart metadata), and results can be inspected with bounding boxes for QA product/docs docs.

Teams can use ProSights through a web playground called Canvas to test files without code, a production API and Python SDK for programmatic use, and an embeddable widget to plug extraction into internal apps. For enterprise deployments, the company offers single‑tenant VPC options, SOC 2 controls, zero data retention, and a commitment not to train on customer data, reflecting a security posture aimed at finance customers Canvas docs security.

Who are their target customer(s)

• Private equity analysts and portfolio-operations teams: They spend hours copying tables and figures from investor decks and portfolio reports into Excel/PowerPoint, which is slow and error‑prone; they need structured outputs they can trust and reuse product/docs.
• Investment‑banking analysts and deal teams: They need accurate extraction from pitchbooks, CIMs, and diligence packs under tight deadlines, with clean exports to Excel/PPT for models and presentations docs.
• Hedge‑fund research and quant teams: They require auditable, programmatic outputs (JSON/SDK) with cell‑level data and chart time series to feed models and backtests, not manual scraping docs.
• Consulting and due‑diligence teams: They process large batches of client PDFs and must quickly turn them into standardized templates and slide decks while preserving an audit trail; a QA‑friendly UI and Excel/PPT exports help Canvas.
• Internal engineering/data‑platform teams at PE, banks, and funds: They need extraction that meets strict security and deployment requirements (single‑tenant, SOC 2, no data retention) and can be embedded into internal pipelines via API/SDK security.

How would they acquire their first 10, 50, and 100 customers

• First 10: Run white‑glove pilots with top PE and bank teams, delivering short, measurable projects (e.g., extract specific decks, generate PPT/Excel) via Canvas and the production API, with single‑tenant/VPC where required Canvas security.
• First 50: Turn early anchors into case studies and sell templated 30‑day pilots to the top PE funds, bulge‑bracket banks, and consultancies. Provide a low‑friction path from free credits to documented exports and an integration playbook to move from manual copy/paste to programmatic ingestion pricing core concepts.
• First 100: Scale self‑service and developer channels (richer SDKs, widget connectors, template libraries) and add partners to handle integrations, while hardening enterprise offerings (SOC 2, no data retention, VPC) to clear procurement and IT reviews product security/docs.

What is the rough total addressable market

Top-down context:

ProSights participates in the intelligent document processing/document AI market, which analysts estimate at roughly $2.3B in 2024 and growing to $12.35B by 2030 (33% CAGR), with finance as a major adopter Grand View Research.

Bottom-up calculation:

Assuming 1,500 finance firms across PE, investment banking, hedge funds, and consulting adopt purpose‑built extraction with average ARR of $25k–$150k per firm (mix of per‑page and enterprise contracts), the core finance TAM is roughly $38M–$225M. Expanding to adjacent BFSI and back‑office use cases would raise this materially.

Assumptions:

• Targetable firms include mid‑to‑large PE funds, banks, hedge funds, and consultancies with recurring document‑extraction needs and compliance constraints.
• Average ARR per firm blends self‑serve/per‑page usage and enterprise contracts with security/deployment add‑ons.
• Focus is on finance‑specific workflows (decks, financials, charts) rather than the entire generic IDP market.

Who are some of their notable competitors

• Amazon Textract: AWS OCR/data‑extraction API for text, tables, and forms at cloud scale; strong if a team is standardized on AWS, but generic out of the box and typically requires more custom integration for finance workflows.
• Google Document AI: Google Cloud’s document‑processing platform with prebuilt processors and custom model tooling; powerful for GCP shops and batch processing, less tailored to finance‑specific slide/report workflows by default.
• Azure AI Document Intelligence (Form Recognizer): Microsoft’s service for extracting layout, key‑value pairs, and tables, integrated with Azure identity and tooling; a fit for Azure‑centric enterprises, but a general document API rather than finance‑workflow‑specific.
• ABBYY (FlexiCapture / Vantage): Enterprise IDP with configurable pipelines and on‑prem/cloud options suited to regulated environments; proven but heavier to deploy and less developer‑first than API/SDK‑centric tools.
• Rossum: IDP platform focused on transactional docs (invoices, POs) with validation UI and connectors; closer on line‑item/table extraction but oriented to AP/ERP use cases rather than investor decks and chart extraction.