Intelligence Vault: Claude x Kensho Integration

Everyone saw the Claude for Finance announcement.

Fewer noticed the Kensho partnership — which more technically shows how GenAI is actually being integrated into financial workflows.

Two weeks ago, Anthropic launched Claude for Financial Services, a tailored version of its LLM for finance professionals.

Alongside it, S&P Global’s Kensho Labs announced a technical integration with Claude that reveals what this means in practice.

Here’s what the integration enables:

Users can ask Claude financial questions and receive answers grounded in live S&P Global data, including:

📈 Capital IQ Financials

📈 Compustat® datasets

📈 Market data

📈 Earnings call transcripts

📈 Business relationships and competitor profiles

These are structured, licensed data sources — not scraped summaries or PDFs. Claude queries specific metrics or insights, then returns grounded answers.

Examples:

💬 “What’s the YoY trend in Tesla’s EBITDA margin?”

💬 “Compare Q2 performance across large-cap semiconductors.”

💬 “Summarise sentiment shifts in Microsoft’s last earnings call.”

✅ Answers include links back to source data, ensuring transparency and auditability — essential in regulated financial environments.

🧩 This is powered by Kensho’s LLMready API, a data access layer built specifically for large language models. Kensho also built a Model Context Protocol (MCP) server to make these datasets queryable inside Claude.

Claude can now:

🔹 Understand structured financial schema

🔹 Retrieve only what’s relevant

🔹 Compose verifiable answers backed by trusted data

🎯 The functional impact:

🔹 Analysts get financial insights without switching tools

🔹 Reporting and peer comparisons become faster, more interactive

🔹 GenAI becomes explainable — with data lineage intact

This might be a blueprint for domain-grounded GenAI in enterprise settings — where data provenance and entitlements matter as much as language generation.

📊 See more details in the image below.