{"slug": "agentic-ai-in-action-part-32-one-name-two-definitions-and-a-coco-skill-that-the", "title": "Agentic AI in Action- Part 32: One Name, Two Definitions, and a CoCo Skill That Picks the Right One", "summary": "A Snowflake practitioner published the 32nd installment of an agentic AI series showing how a single CoCo skill file, SKILL.md, makes the assistant ask users which team's definition of net_revenue they mean before querying the matching governed semantic view, since Finance's REVENUE_SV on FINANCE.INVOICES and Sales' BOOKINGS_SV on SALES.BOOKINGS produce figures that differ by 13,700 on the sample data. The approach relies on Snowflake's Business Ontology, which the author notes is now in private preview, while the post uses capabilities CoCo already supports. The author also announced being named Winner in the AI Excellence category for APJ at the Snowflake Community Awards 2026.", "body_md": "*Before we begin, on a personal note, I have some news to share. I was named the Winner in the* *AI Excellence* *category for APJ at the Snowflake Community Awards 2026. The work behind it includes writing my book, Ultimate Snowflake Cortex AI for Generative AI Applications, publishing an agentic series of blogs with code on GitHub, building multiple marketplace apps, winning a hackathon, and helping build our internal practice. This award recognizes that work, but it also belongs to the people who read the posts, tried the code and wrote back with questions, because their feedback shaped much of what I wrote next.* *Thank you* *to everyone who nominated, voted, read and shared! Checkout the announcement* *here* *and the awards video* *here.*\n\nNow, on to the post.\n\nAsk for net revenue and the right answer depends on who is asking. Finance means *recognized revenue after refunds and discounts*, with intercompany invoices and test accounts removed. Sales means *bookings net of discounts given at signing*. Each team has a governed semantic view with a metric called net_revenue, and both are correct. An agent with no guidance will pick one and not tell you which.\n\nSnowflake’s Business Ontology is aimed at this problem, and is now in private preview. This post uses what CoCo already supports today. Two semantic views hold the numbers, and a single skill file tells CoCo which definition to use and how to get the value.\n\nThe scenario is a business user asking CoCo, in plain language, what net revenue was.\n\nFinance and Sales each report it from different tables and with different rules, and each team already has a governed semantic view for its version. The goal is for CoCo to ask which team’s definition the user means, and then answer from that team’s view and say where the number came from.\n\nStart with a database, one schema for each team, a small warehouse, and a table for each team.\n\n*FINANCE.INVOICES* *is what Finance billed, with one row per invoice and the refunds and discounts recorded against it:*\n\n*SALES.BOOKINGS* *is what Sales closed, with one row per signed deal and the discount given at signing. The rows are fixed, so the numbers later in the post are easy to check.*\n\nFinance works from invoices, which carry refunds and discounts, plus one intercompany row and one test account row that should never count as revenue. Sales works from bookings, which carry the discount given at signing. Those differences are why the same name will produce two different numbers.\n\nWith the data in place, define one semantic view for each team, and give both a metric called net_revenue.\n\nREVENUE_SV sits on INVOICES and holds Finance’s definition, which subtracts refunds and discounts and leaves out the intercompany and test account rows.\n\nBOOKINGS_SV sits on BOOKINGS and holds Sales’ definition, which is the booking amount minus the discount given at signing.\n\nNeither view is wrong. They answer different questions under the same name.\n\nWith the sample data the two numbers differ by 13,700. A user who asks CoCo for net revenue gets one of them, depending on which view or table it happens to reach for.\n\nA skill is a markdown file of instructions that CoCo reads and follows when a request matches it. In this case it gives CoCo the two approved definitions of net revenue and the query for each, so CoCo asks which one you mean and answers from the governed semantic view.\n\nSave the skill file from the repo on your computer as SKILL.md, then add it from the CoCo window in Snowsight.\n\n3. Type / in the message box and check that net-revenue-definitions appears in the list.\n\nThe skill file content is shown below:\n\nThe skill file has four parts. The front matter gives the skill a name and a description, and CoCo matches the description against your prompt to decide when to load it. The table records the two approved definitions and the semantic view behind each one. The workflow tells CoCo to ask which team you mean, run the matching SEMANTIC_VIEW query, and answer with the number, the definition and its source. The rules stop it from calculating net revenue from the raw tables. The full file in the repo also covers breakdowns by region and month.\n\nIn the same workspace, open CoCo and ask what net revenue was.\n\n“What was the net revenue”\n\nCoCo asked which definition I meant, ran a SEMANTIC_VIEW query, and returned 48,100 labelled governed, which is the Finance team’s number to look for with the sample data. Answering Sales should return 61,800 from the bookings view.\n\nWhen one business term has two meanings, an agent needs both of them written down and a rule for choosing between them. Two semantic views hold the governed numbers, and a single skill file records both definitions, tells CoCo to ask which one you mean, and points it at the right view.\n\nThe answer then comes from a governed metric and not from a calculation CoCo improvised. Keeping the definition next to the query also means anyone reading the answer can see what the number means as well as what it is. Business Ontology is the managed version of this idea, with owners, approvals and an audit trail, and a skill like this one is a simple way to start.\n\nThe setup script and the skill file for this blog can be accessed [here.](https://github.com/Krishsriniv/coco-net-revenue-definitions-skill)\n\n*I share hands-on, implementation-focused perspectives on Generative & Agentic AI, LLMs, Snowflake and Cortex AI, translating advanced capabilities into practical, real-world analytics use cases. Do follow me on* *LinkedIn* *and* *Medium* *for more such insights.*\n\n[Agentic AI in Action- Part 32: One Name, Two Definitions, and a CoCo Skill That Picks the Right One](https://pub.towardsai.net/agentic-ai-in-action-part-32-one-name-two-definitions-and-a-coco-skill-that-picks-the-right-one-4b38db5a6077) was originally published in [Towards AI](https://pub.towardsai.net) on Medium, where people are continuing the conversation by highlighting and responding to this story.", "url": "https://wpnews.pro/news/agentic-ai-in-action-part-32-one-name-two-definitions-and-a-coco-skill-that-the", "canonical_source": "https://pub.towardsai.net/agentic-ai-in-action-part-32-one-name-two-definitions-and-a-coco-skill-that-picks-the-right-one-4b38db5a6077?source=rss----98111c9905da---4", "published_at": "2026-11-10 05:05:17+00:00", "updated_at": "2026-11-10 05:22:58.604247+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "artificial-intelligence"], "entities": ["Snowflake", "CoCo", "Snowflake Business Ontology", "Snowflake Community Awards 2026", "FINANCE.INVOICES", "SALES.BOOKINGS"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/agentic-ai-in-action-part-32-one-name-two-definitions-and-a-coco-skill-that-the", "markdown": "https://wpnews.pro/news/agentic-ai-in-action-part-32-one-name-two-definitions-and-a-coco-skill-that-the.md", "text": "https://wpnews.pro/news/agentic-ai-in-action-part-32-one-name-two-definitions-and-a-coco-skill-that-the.txt", "jsonld": "https://wpnews.pro/news/agentic-ai-in-action-part-32-one-name-two-definitions-and-a-coco-skill-that-the.jsonld"}}