FinSight | Custom Datasets for Portfolio Managers

At a glance

Product: FinSight, self-service dataset intelligence by ForeStrat
What it does: Converts documents and feeds into structured, ontology-mapped datasets for portfolio managers and research workflows
Pipeline: Ideate → Ingest → Extract → Validate → Review (optional) → Publish
Sources: Public filings, private documents, scanned files, web scraping, email, SFTP, API, social media feeds
Heritage: Built by the team that created ForeSeer, acquired by S&P Global in 2022; extraction technology used across S&P Market Intelligence
Platform: Deployed on DataStrat inside your secure infrastructure
Demo: Book a demo at forestrat.ai/about.html#contact

Proven extraction pedigree

From ForeSeer to FinSight

FinSight was built by the same team that created ForeSeer, the generative AI data platform acquired by S&P Global in 2022. ForeSeer technology powers information extraction across S&P Market Intelligence today, spanning company fundamentals, consensus estimates, market data, people data, and more.

Fundamentals

Structured financials and filing-derived metrics at scale across global issuers.

Estimates

Consensus and broker estimate extraction from research and corporate disclosures.

Market data

Reference, pricing, and instrument-linked attributes pulled from unstructured sources.

People data

Executive and leadership intelligence mapped from filings, bios, and corporate documents.

FinSight carries that same extraction DNA forward, now as a self-service platform your team runs inside your own environment.

Any source. Structured output.

Whether you're extracting nuanced fundamentals from global filings, trade ideas from broker research, or sector metrics from scraped web pages, FinSight gives your team a self-service path from ideation to governed production without one-off scripts or manual spreadsheets.

Public documents

SEC EDGAR and global filing coverage across US, EU, APAC across all developed markets, all major emerging markets, and all major commodity markets, multi-lingual support with auto-translation.

Private documents

Broker research PDFs, internal memos, and email attachments via IMAP mailbox, SFTP, or direct upload. Secured credential vaulting to ensure data privacy and security.

Web scraping

Scrape content from free running text or tabular data from various sources including IR pages, rail reports, commodity sites, and sector portals. Scheduled crawling with change detection and headless browser support for JS-heavy sites.

From documents to production datasets

FinSight replaces ad-hoc extraction scripts with a governed, repeatable pipeline your team controls end to end, backed by ForeStrat's institutional data platform.

Self-Service Ideation

Define what you want to extract, examples include but not limited to fundamentals, trade ideas, rail carloads, sector metrics, with a versioned ontology and testable prompts. Go from concept to pilot without an engineering ticket. Guided workflows help you sample against real documents, approve your schema, and productionize the pipeline.

Universal Ingestion

One platform for every source type. Pull from public registries, private mailboxes, SFTP drops, REST APIs, and the open web. Per-client channel configs, credential vaulting, content-hash deduplication, and scheduled cadence keep ingestion reliable at scale.

AI-Powered Extraction

Multi-model extraction with adaptive routing. Deterministic parsers where structure is reliable, LLMS where it isn't. Every value maps to a canonical ontology with full source provenance: page, table, row, and bounding box. Built for finance. Think fundamentals, broker trade ideas, and custom sector datasets alike.

Data Quality Checks

FinSight agents run automated checks before datasets reach production: error detection, tally and balance validations, cross-field consistency, and period-over-period sanity checks. Configure client-specific rule packs for your funds, sectors, and ontology. Exceptions route to the right owner with full context, so bad data never ships downstream.

Analyst Review & Quality Loop

Split-pane review UI puts the source document alongside structured fields. Click to highlight in source on every value. Ground-truth regression for public filings; human approve/reject for private research. Corrections feed back into future extractions, compounding accuracy over time.

Publication & Delivery

Deliver approved datasets to portfolio managers via watchlists, streaming APIs, and digests. Export structured output as JSON, CSV, or Parquet for downstream models and research workflows. FinSight closes the loop from raw document to consumable intelligence.

Built for Finance

FinSight is finance focused. Once the ontology is defined any domain dataset is supported on FinSight, be it broker research and trade ideas, custom fundamentals, sector alt data, or some bespoke fully custom private datasets.

Example Datasets

Trade ideas

Broker research PDFs and email → structured ideas with instrument, direction, conviction, and rationale.

Public filings

SEC 10-K/10-Q and global registries → ontology-mapped financials with XBRL alignment.

Sector / alt data

Scraped IR pages, rail reports, commodity sites → scheduled metrics with dated periods.

Custom datasets

Any document type you define → your ontology, your schema, your production pipeline.

Built for your whole team

Ontologist / Data Engineer

Defines the extraction ontology, prompts, and schema. Versions and publishes bundles that govern what the pipeline extracts.

Analyst

Reviews extractions in a split-pane UI, approves or rejects fields, and improves quality with every correction.

Portfolio Manager

Consumes published trade ideas and structured research via watchlists, digests, and AI ready data via API. Interpret manually or feed to models.

FinSight is built on ForeStrat's DataStrat platform, the same infrastructure that powers institutional data pipelines at scale.

Frequently Asked Questions

What is FinSight? +

FinSight is ForeStrat's self-service dataset intelligence product. It turns public filings, private research, unstructured documents, social media feeds, and web-scraped content into structured, ontology-mapped datasets, with multi-channel ingestion, AI-powered extraction, optional human-in-the-loop review, and delivery to portfolio managers.

What is the relationship between FinSight and ForeSeer? +

FinSight was built by the same team that created ForeSeer, the generative AI data platform acquired by S&P Global in 2022. That extraction technology still powers information extraction across S&P Market Intelligence today, spanning fundamentals, estimates, market data, people data, and more. FinSight carries that pedigree forward as a self-service platform deployed in your environment.

What sources can FinSight ingest? +

FinSight ingests public filings and global regulatory documents, private broker research and email attachments, scanned documents, web-scraped content, SFTP drops, REST APIs, and social media feeds. Channels support scheduled cadence, credential vaulting, and content-hash deduplication.

What datasets can FinSight produce? +

Example datasets include structured trade ideas from broker research, ontology-mapped financials from SEC and global filings, sector and alternative data metrics from scraped sources, and fully custom private datasets defined by your team's ontology and schema.

Who is FinSight built for? +

FinSight serves ontologists and data engineers who define extraction schemas, analysts who review and approve extractions, and portfolio managers who consume published datasets via watchlists, digests, and API exports for research and models.

How does FinSight handle data quality? +

FinSight runs agent-driven quality checks on every dataset before publication: error detection, tally and balance validations, cross-field consistency, and period-over-period sanity checks. You can define client-specific rule packs tailored to your funds, sectors, and ontology. Failed checks surface as routed exceptions with full provenance, so teams fix issues before data reaches portfolio managers or models.

How do I get a FinSight demo? +

Book a demo via our contact form. ForeStrat deploys FinSight on top of the DataStrat platform inside your secure infrastructure.

Ready to turn documents into Alpha?

Book a demo to see how FinSight can power your research and alt-data workflows

Book Demo