Snowflake Intelligence vs Cortex AI: The Real Product Boundary (mid-2026)
The question
What is the actual product boundary and feature split between Snowflake Intelligence and Cortex AI as of mid-2026, and — for a phData Deal Solutions Architect — where does a CAF/Brigade deployment plug into that stack? phData's project index names BOTH as targets, but no vault doc maps their real feature boundaries, so every discovery conversation risks imprecise framing. The right placement matters for the phData-Snowflake pitch: the DSA has to say cleanly which capability lives where.
What we already know (from the vault)
- Snowflake Intelligence is the natural-language UI on top of the Cortex stack. GA November 2025; lets non-technical users query structured + unstructured data in English — the "ask anything" front door to Cortex. See [[2026-04-28-snowflake-enterprise-knowledge-management]].
- Cortex Agents (GA Nov 2025) are the orchestration layer. A plan-execute-respond loop that uses Cortex Search (hybrid retrieval over unstructured text) and Cortex Analyst (managed text-to-SQL over structured data, backed by a semantic model YAML) as tools. Cortex AISQL / AI_* functions (AI_COMPLETE, AI_EXTRACT, AI_PARSE_DOCUMENT, AI_EMBED, etc.) are the primitive LLM operations underneath. See [[2026-04-28-snowflake-enterprise-knowledge-management]] and [[study-cortex-ai]].
- The certification material treats them as distinct privilege objects. Analyst needs access to the semantic view/model; Search needs privileges on the search-service object; Agents combine Search + Analyst; Snowflake Intelligence is its own layer with its own grants. This is a strong tell that Snowflake itself models SI and the Cortex services as separate objects, not one product. See [[study-cortex-ai]].
- phData's delivery spine builds at the Cortex Agents layer, not the SI-UX layer. The skills-first workflow is
setup → create-views → create-semantic-view → create-agent → deploywrapping Cortex Search + Cortex Analyst + Cortex Agents with phData IP (golden-set / TruLens eval). CAF is the assessment-to-build engine that decides which agents to build and emits a build manifest; it spans the sales→delivery seam and sits above the whole stack as method. See [[2026-06-28-snowflake-si-cortex-positioning-caf-gap]]. - Brigade can run inside Cortex Code (now CoCo). CoCo ships a Claude-Code-compatible plugin/skill/subagent model, so the phData skill-agent-brigade (spec → test → code → critic) can execute within Snowflake's boundary via CoCo Cloud Agents — which is what defuses the external-Anthropic-egress objection. See [[2026-07-05-brigade-skills-snowflake-cortex-code]].
What the web says
The boundary is real and consistent across sources — but the names changed at Snowflake Summit 2026 (June 2, 2026), which is the single most important thing for a mid-2026 conversation:
- The layer split is clean and Snowflake-confirmed. "Snowflake Intelligence is essentially Cortex Agents + Cortex Analyst + Cortex Search + Cortex AI Functions all working together in a secure Snowflake environment. Snowflake Intelligence is the user-facing interface and application layer, while Cortex Agents, Analyst, and Search are the technical building blocks." (WebSearch synthesis of Snowflake + dbt Labs + Flexera coverage.)
- The "Agentic AI Platform" is the umbrella, and SI is one member of it. Snowflake now frames the whole thing as one platform "consisting of Snowflake CoWork, Cortex Agents, Cortex Search, Cortex Analyst, Cortex AISQL, Semantic Models, Openflow, and Horizon governance working together." Note SI/CoWork is listed inside the Cortex AI suite — it is the top, user-facing product of the family, not a rival to it. (Atlan Summit recap)
- RENAME 1 — Snowflake Intelligence → Snowflake CoWork. Now positioned beyond Q&A as "the personal agent for every knowledge worker" and "the system of action" — it takes action (Artifacts, PDF/PPT generation, User Skills, User Memory, Deep Research, MCP connectors to Slack/Jira/Gmail/Salesforce, iOS app). GA at Summit. (Snowflake CoWork press release)
- RENAME 2 — Cortex Code → Snowflake CoCo. The developer-facing agent surface (coding agent in desktop app, Snowsight, terminal, IDEs). Launched Feb 2026; fastest-growing product in Snowflake history (~7,100 accounts by Summit). CoWork is the business-user agent; CoCo is the developer agent — two sibling app surfaces over the same Cortex services. (Summit 2026 feature summary)
- NEW — Cortex Sense. A shared context / enterprise-memory layer built from query history, metadata, dashboards, and business definitions; feeds both CoWork and CoCo. Snowflake claims "83% accuracy vs 24% without it." This slots between the agent surfaces and the retrieval services. (Summit 2026 feature summary)
- Cortex Agents is the hinge. It "reasons over a request, plans the work, calls tools, runs code, and returns an answer" (GA Nov 4, 2025) — the orchestration engine that both the packaged CoWork UX and any custom-built agent app sit on top of, and it is exposed via REST API for custom apps. (Cortex Agents docs)
Convergences and contradictions
- Convergence: Vault and web agree on the same stack shape — SI/CoWork = user-facing agentic app; Cortex Agents = orchestration; Cortex Analyst (structured) + Cortex Search (unstructured) = the two tools it orchestrates; AISQL / AI_* = primitive LLM functions; Horizon = governance; semantic models = the business-definition bridge. The cert material's separate-privilege-object framing directly corroborates the web's "app layer vs building blocks" split.
- Contradiction / sharpening #1 — it is NOT "SI vs Cortex." The intuitive framing (two competing products) is wrong. "Cortex AI" is Snowflake's umbrella brand for the entire AI suite, and Snowflake Intelligence / CoWork is one product within it. The correct axis is app/UX layer (CoWork + CoCo) vs services layer (Agents, Analyst, Search, AISQL, Sense, Semantic Models) — not SI against Cortex. A DSA who frames them as alternatives will sound wrong to a Snowflake SE.
- Contradiction / sharpening #2 — the names shifted at Summit 2026 and the boundary moved "up." SI was answer-questions in Nov 2025; as CoWork it is now action-taking (generates artifacts, runs skills, holds memory, reaches out via MCP). Third-party docs, RFPs, and even some Snowflake pages still say "Snowflake Intelligence" and "Cortex Code." Expect mixed naming through 2026. Speaking precisely in mid-2026 means naming both: "Snowflake Intelligence, now CoWork."
- Honest ambiguity: the newest pieces (Cortex Sense, Cortex Training) are preview/early-GA and blur the once-clean two-layer picture into more of a three-tier stack (surfaces → context → services). The core boundary is stable; the mid-tier is still hardening.
Synthesis for RDCO
The crisp boundary statement. Snowflake Intelligence (rebranded Snowflake CoWork at Summit 2026) is the agentic application / UX layer — the conversational, action-taking front door a business user opens and talks to. Cortex AI is the umbrella suite of underlying services that CoWork (and any custom app) is built on: Cortex Agents (orchestration — plan/tool-call/respond), Cortex Analyst (structured, text-to-SQL over a semantic model), Cortex Search (unstructured, hybrid retrieval), Cortex AISQL / AI_ functions* (primitive LLM ops), plus the newer Cortex Sense (shared context) and the Semantic Models + Horizon governance substrate. The mental model: CoWork/CoCo = the destination app; Cortex AI = the engine room; Cortex Agents = the hinge between them. CoWork is the business-user surface, CoCo (formerly Cortex Code) is the developer surface — two apps over the same services.
How a DSA should frame it in discovery. When a client says "we want Snowflake Intelligence," first translate the name ("that's now called CoWork") and then move the conversation one layer down, because SI/CoWork itself is essentially a toggle — the real engineering is in Cortex. The value phData delivers is not "turn on CoWork"; it is making the layer underneath trustworthy and governed: modeling the semantic view so Cortex Analyst answers structured questions correctly, standing up Cortex Search services over the client's unstructured corpus (with RBAC-aware, metadata-filtered retrieval), and wiring a Cortex Agent that orchestrates both with explicit tool boundaries and eval. A phData-built agent can then be surfaced through CoWork for business users, or through a custom Streamlit/MCP app — same Cortex plumbing, different face. The line to hold: SI/CoWork is what the client's users touch; Cortex Agents + Analyst + Search + AISQL is what phData builds and governs.
Where CAF/Brigade attaches — exactly. CAF (Catalyst Assessment Framework) sits above the entire stack as method: it assesses the estate, classifies use cases, routes by autonomy, and emits a build manifest — it decides which Cortex Agents to build before anyone builds one. It attaches at the front of the funnel (discovery/scoping), not to any single product box. Brigade (the four-station spec→test→code→critic skill-agent) is the factory that produces the build artifacts CAF specifies, and it can execute inside Snowflake's boundary via CoCo (Cortex Code) because CoCo ships a Claude-Code-compatible plugin/skill/subagent model — so the delivery work lands at the Cortex Agents + Analyst + Search layer, in-perimeter, avoiding external Anthropic egress. Net: CAF = the method above the stack; Brigade = the in-CoCo build engine; the deliverable = a governed Cortex Agent; CoWork = the optional business-user face on top.
Open follow-ups
- Does Cortex Sense change the phData engagement scope? If Snowflake now auto-assembles the context layer (query history, metadata, definitions), how much of the "model the semantic layer once" wedge does it absorb vs. leave for phData?
- Post-rename, is "Snowflake Intelligence" being fully retired or kept as a doc alias? Confirm which name Snowflake SEs and RFPs will standardize on for H2 2026 so the DSA uses the client-current term.
- Where does the packaged-plugin motion (CoWork finance/sales prebuilt plugins) collide with CAF's custom assessment-to-build engine — is that Snowflake competing with the partner delivery layer at mid-market?
- Can a phData-built Cortex Agent be exposed natively as a CoWork skill/plugin, or only via custom app / MCP? This determines whether phData deliverables inherit CoWork's UX for free.
- What exactly is GA vs preview across CoWork Skills/Memory, Cortex Sense, and Cortex Training as of the client's contract date — the Summit "GA" claims and the press-release "preview soon" language disagree in places.
Related
- [[2026-06-28-snowflake-si-cortex-positioning-caf-gap]] — big-SI Cortex positioning and where CAF's mid-market wedge sits; the phData delivery spine this brief builds on
- [[2026-04-28-snowflake-enterprise-knowledge-management]] — Snowflake's KM stack; the "SI is the NL UI on top of Cortex Search + Cortex Agents" source
- [[study-cortex-ai]] — SnowPro Gen AI C02 notes: Cortex Analyst / Search / Agents function-level detail and the separate-privilege-object framing
- [[2026-07-05-brigade-skills-snowflake-cortex-code]] — Brigade runs inside CoCo (Cortex Code); the in-perimeter build path for CAF deliverables
- [[2026-05-20-phdata-cortex-agents-practice]] — phData's Cortex delivery spine, deal shape, and where the build artifacts land
Sources
Vault:
- [[2026-06-28-snowflake-si-cortex-positioning-caf-gap]] —
~/rdco-vault/06-reference/research/2026-06-28-snowflake-si-cortex-positioning-caf-gap.md - [[2026-04-28-snowflake-enterprise-knowledge-management]] —
~/rdco-vault/06-reference/research/2026-04-28-snowflake-enterprise-knowledge-management.md - [[study-cortex-ai]] —
~/rdco-vault/01-projects/certifications/snowpro-genai-c02/study-cortex-ai.md - [[2026-07-05-brigade-skills-snowflake-cortex-code]] —
~/rdco-vault/06-reference/research/2026-07-05-brigade-skills-snowflake-cortex-code.md - [[2026-05-20-phdata-cortex-agents-practice]] —
~/rdco-vault/06-reference/research/2026-05-20-phdata-cortex-agents-practice.md
Web:
- Snowflake CoWork press release (SI → CoWork rename, app-layer definition): https://www.snowflake.com/en/news/press-releases/snowflake-cowork-powers-the-agentic-enterprise-as-the-personal-agent-for-knowledge-workers-to-work-smarter/
- Snowflake Summit 2026 feature summary (Umesh Patel, Medium) — full component list + GA status: https://medium.com/snowflake/snowflake-summit-2026-summary-of-new-features-09f3d5ffeefe
- Atlan — Snowflake Summit 2026 announcements (Agentic AI Platform composition): https://atlan.com/know/snowflake/summit-2026-announcements/
- Snowflake "control plane for the agentic enterprise" release (SI + Cortex Code expansion): https://www.snowflake.com/en/news/press-releases/snowflake-expands-snowflake-intelligence-and-cortex-code-to-power-the-control-plane-for-the-agentic-enterprise/
- Cortex Agents — Snowflake docs (orchestration definition, GA date): https://docs.snowflake.com/en/user-guide/snowflake-cortex/cortex-agents
- Snowflake Cortex AI product page (umbrella suite framing): https://www.snowflake.com/en/product/features/cortex/
- Flexera — Snowflake Intelligence 101 (2026 overview): https://www.flexera.com/blog/finops/snowflake-intelligence/