Building an AI agent is easier when you already have a BI foundation to draw from. Traditional BI already carries some context, data connections, and curated understanding that took me years to build. The hard part to produce meaningful conversational analytics is reproducing that context in an agent consumable format, with enough accuracy and consistency to trust it. Once you clear that bar, the real value reveals itself, with new challenges. Answer questions BI never answered or lose momentum, excitement, and trust. Agent lifecycle management is a CI/CD and governance exercise, not a BI delivery exercise, and that difference is what breaks me through the historically high BI project failure rate.
An Agile Foundation
BI architects build durable systems meant to answer a wide range of questions, including ones not yet asked. Semantic models give builders, and now LLMs, coverage, context, and a declarative path into governed data. All of that work still matters… But understanding business at the speed business happens requires aglity and speed. A governed structure dooes not always pull in the same direction. My setup: a mono-repo holding the dbt models, the semantic models, and the resources that feed skills, plus a metrics glossary deployed independently for active development.
Framing Real Conversations in Conversational Analytics
The conversation between a user and an agent should stay a private, user-controlled space for exploration, ideation, and iteration. Users need room to ask questions, test ideas, bring in their own data, and work through half-formed or sensitive thinking, without assuming the full history becomes visible, or worse, becomes context for someone else’s session. For executive users this is a governance problem you know, especially at a public company. That conversation history can carry strategic initiatives, partnership terms, trade secrets, and other need-to-know risks.
That underling data and resulting analysis work still has to follow governance policy. Access to it matters as much as access to the underlying data itself. Too little attention goes to what sits inside “context” and the conversation logs it produces.
Participation and training users to make effective use of this powerful conversational analytics tool cannot be under-stated. To get access to Claude + Snowflake for me, requires a mandatory 1 hour lab session. The impact and influence on decisions drives the population who get access. This observation period is how everyone in the room learns and how I designed my feedback loops to scale this process.
Instead of treating the conversation itself as feedback, the Chat Feedback Skill closes these loops automatically. It helps the user build a deliberate representation of the session and submit it on purpose. The agent’s Chat Feedback Tool receives that submission and processes it as feedback. Keeping the private workspace separate from the intentional feedback artifact preserves psychological safety and governance. It still gives users a direct, structured way to help improve the agent.
Closing the Decision Loop: Embedded AI BI Where Conversations Occur and Work Happens
Adoption comes down to one thing: embed analytics where people already work. Logging into a BI portal is a vendor-prescribed experience, and at best I’d call it barely tolerable. At organizations that have adopted Claude for Business, I have watched the way of working actually shift. That will keep evolving, and Claude will likely end up a stepping stone toward whatever comes next, not the final destination.
Semantics over Semantics
Plenty of prescribed arrangements apply semantics to data and context for interpretation and inference. But even the perfect arrangement, right information to the right person, is not enough on its own. The loop looks like this:
- Information seeking requires data and analysis.
- Data and analysis create understanding.
- Understanding requires more information to act on.
- An artifact explains the new information and shares that understanding.
- Communication and authority carries the directive and the action.
Claude lets decision makers move faster through this loop, with thin semantics and narrowly scoped data access. That’s the standard conversational analytics demo. The problem shows up after the demo ends: the action lives in memory, and the data platform stays disconnected from it. There are many approaches to help integrate these decision loops with descriptive statistics in a LLM powered experience. Vertical integration with embedding agents inside of Slack and Teams is one tactic that I think will play a big part to activate these decisions.
Snowflake Semantic Layer
My early success didn’t require a semantic layer, mostly because I had built very strong context artifacts. I benefit as a BI practitioner who was obsessed already with documentation. Once I built my Snowflake semantic layer into my CI/CD process, the speed and progress compounded.
Closing the BI on AI Data, Semantics, and Context Loop in Snowflake
Instead of framing every component in my process, here’s a schematic of my context loops: This is the shape of my AI BI loops heading into Q4 2026. It will look different again by Q1 2027 as Snowflake, Claude, and my own experience absorb more of the knowledge work. This covers two loops I run to keep the system improving. It is not the full set of loops, and neither loop is fully closed. The DataTools metrics glossary runs across every phase and plane as a universal context layer.
Understanding Adoption and Utilization: The Data Loop with DataTools Radar
Snowflake ships a strong set of observability tools. Observability gives a structured, narrowly scoped replay of what happened inside Snowflake. That raw data carries real value, but not the direction needed to actually improve the conversational analytics experience. Snowflake’s foundation is flexible enough to build a bespoke layer on top, tuned and contextual. To turn that into actionable intelligence, I built this DataTools Radar. This first figure gives me an over-arching catalogue of my Snowflake estate based on questions I need answers to.
Automating Recommendations and Remediation
Understanding what happened is nice. Acting on it immediately is the objective function. I measure success by adoption and session utilization. Alongside Snowflake observability, a set of analyst skills pushes real conversational context back into Snowflake through a stored Snowflake stored procedure (added as a tool to my agent).
AI earns its keep here: sifting that context and surfacing recommendations is what lets the right decision makers work the right problems. The “Profiled” questions come straight from real conversations, grouped into 48 recurring topics. That is productized proactivity, something Business Intelligence rarely delivered.
Business Conversational Loop
In my multi-agent setup, the business conversation happens in Claude, because that’s where the work already happens. My job is protecting that flow: masking and filtering anything moving from Snowflake into Claude that shouldn’t leave Snowflake unmasked. I also need real context and knowledge artifacts captured as they’re created. For issues and continuous improvement, a feedback skill redacts and structures the session into a clear feedback loop for remediation.
The output gives the analytics engineer actionable guidance directly. 8 of 10 of these feedback loops typically go into Claude Code with no adjustment before landing as an approved pull request.
What’s Next?
These loops are continuously evolving and improving alongside my ability and understanding how to evaluate the efficacy and accuracy of the agents themself. Today, the Snowflake Cortex agents hold very little context outside of the semantic layer. As I get shaper tooling and results building evaluations to find material improvements that will shift into 2027.
