Cases

Financial agent on MCP, SQL, and executable skills

How KT.Team moved OSNO-VA's financial logic from spreadsheets into MCP, SQL, and executable skills for AI agents.

Our clients

Clients and partners

Capital Group
FSK Group
SMLT
Tochno
Dogma
Sber City
FM Logistic
Danone
Relief Center
Pandora
OSNO-VA case - financial agent on MCP
Saint-Gobain
Askona
FIX PRICE
Snezhnaia Koroleva
Muztorg
TVOE
Greenway
Polaris
Campari
Yandex
Lenta
International perfume and cosmetics brand
OSNO-VA case - financial agent on MCP
RAEC
EKF
L'Etoile
Inventive Retail Group

Key takeaways

  • How KT.Team moved OSNO-VA's financial logic from spreadsheets into MCP, SQL, and executable skills for AI agents.
  • Delivered by KT.Team. The CIS source page carries the full project story, metrics and interface screenshots.

Challenge

Cases Financial agent on MCP, SQL, and executable skills How KT.Team moved OSNO-VA’s financial logic from spreadsheets to MCP, SQL, and executable skills for AI agents. 2026-06-04 Discuss the project Key takeaways How KT.Team moved OSNO-VA’s financial logic from spreadsheets to MCP, SQL, and executable skills for AI agents. OSNO-VA needed to give AI agents reliable access to financial data and calculation logic.

This approach lets users maintain financial logic without constant engineer involvement, while the agent can perform checks and calculations under reproducible rules. Task OSNO-VA needed to give AI agents reliable access to financial data and calculation logic. The spreadsheet approach proved too fragile for an agentic scenario: limits, retries, and brittle logic hurt speed and reliability.

OSNO-VA case - financial agent on MCP
Financial AI agent receives data via MCP

Solution

KT.Team separated the general-purpose data platform and the organization’s financial skills: data is translated into SQL/databases, the agent gets access through MCP, and calculation business logic is moved into executable skills and scripts. Financial AI agent gets data through MCP Solution This approach lets users maintain financial logic without constant engineer involvement, while the agent can perform checks and calculations under reproducible rules.

Result - an AI-native financial stack where the agent does not just read a spreadsheet, but calls verifiable tools, gets data through MCP, and works with governed calculation logic.

Results

Blog Topics All articles 23.7.2026 AI accountant for a chain: how to automate revenue reconciliation 23.7.2026 How much an AI agent costs: what drives price and TCO 16.7.2026 AI for accountants: what it can do in 2026, whether it can replace them, and how to use it 27.6.2026 LLM capabilities in 2026: what to choose for your process and budget 19.6.2026 DORA 2026: AI does not deliver ROI - your process does

Platform Projects Related cases All cases OSNO-VA: an AI accountant that performs accounting operations according to rules Built an AI platform #1c #ai #bi #consulting #cost 2026-06-18 Learn more Management accounting OSNO-VA: P&L dashboards in Apache Superset #bi #consulting #integration #superset Learn more Case tags Clicking a tag opens the case catalog filtered by it #ai #bi #consulting #integration #superset All cases

Discuss a similar task: Financial agent on MCP, SQL and…

Enter your email or phone number so we can get back to you.

Send via: