How Cortex Code Is Helping FP&A Move from Reporting to Insight
Finance needs faster answers, not just better reports
For most of my career in FP&A, the work started after the numbers were closed. We would gather data, reconcile it, explain it, package it and then deliver it to the business. The problem was that by the time the answer was ready, the business had often already moved on to the next question.
That changed for me when finance became more deeply connected to our data platform.
I have been using Snowflake for finance for eight years. One of the clearest signals that we were building a different kind of FP&A organization came early: the fourth member of my team was a data scientist. That was intentional. I believed finance needed to move beyond reporting what happened and become a true operating partner to the business. To do that, finance data could not stay trapped inside spreadsheets, static decks and month-end processes. It had to be activated inside the workflows where decisions were actually being made.
That shift has been transformational for my career. Once finance data became easier to access, model and operationalize, my team spent less time acting as a reporting function and more time helping business leaders make decisions. The automation we could achieve also helped us scale that support efficiently.
Why Snowflake became the center of our finance processes
Today, Snowflake sits at the heart of finance’s processes. For many finance organizations, that can sound intimidating. Not every finance team has engineers sitting inside the function, and not every analyst wants to become one. That is one reason I moved to Streamlit in Snowflake years ago. Streamlit gave us a way to turn finance logic into connected, usable applications with a visual interface, while keeping the experience close to the data and under Snowflake’s existing governance model.
What Cortex Code changes is who can build those applications, and how fast.
What Cortex Code changes for finance
Cortex Code is Snowflake’s AI coding agent, integrated directly into Snowflake and designed to understand Snowflake roles, schemas and best practices. In Snowsight, it can help generate, modify, optimize and explain SQL and Python, work with the active workspace as context, and support governance, security and cost-management workflows. It is now generally available in Snowsight for commercial accounts. Snowflake also offers a Cortex Code command line interface, which extends these capabilities beyond the browser and gives teams a flexible way to use the agent in local development workflows while still benefiting from Snowflake-native context and controls.
For a finance organization, that matters because the best AI outputs do not come from prompts alone. They come from prompts plus context. And in enterprise finance, data is context.
That is the real reason to use Cortex Code in Snowflake instead of treating AI coding as a generic external tool. This is not just a model comparison. The real advantage comes from the environment, execution path and data-native context.
That distinction is especially important in finance.
How our team is using Cortex Code
Our team is not using Cortex Code to write toy demos. We are using it to accelerate work that sits directly on top of governed enterprise data. We use it to build connected dashboards in Streamlit. We use it to automate process flows like investigation and outreach. We use it to code applications, including one that automates variance analysis and another that supports long-range planning. Because Cortex Code works inside the same environment where our finance data, permissions and objects already exist, it dramatically reduces the time between a business question and a usable answer.
For me, that is the headline: Cortex Code is helping shrink finance’s time to insight as close to real time as possible.
Why this lowers the barrier to building
It also lowers the barrier to entry. Historically, building finance applications required a heavier technical lift. Even when the logic was straightforward, the plumbing was not. Analysts needed help wiring together data access, interfaces, roles and deployment. Cortex Code changes that. It makes it much easier for non-engineering teams to move from idea to connected application without losing trust or control.
Why governance and security matter so much in finance
There is also a governance and security argument here that matters just as much as productivity.
This is another reason I would position Cortex Code differently than a generic coding assistant. A generic assistant may be very strong at writing code in the abstract. But finance does not operate in the abstract. Finance operates on governed data, under permissions, with auditability concerns, and with a constant need to move from analysis into action. Cortex Code is designed for that environment.
Why the commercial model fits finance
The commercial model also fits how many finance leaders think about technology adoption. Snowflake’s platform for enterprises uses a consumption-based pricing model, and Cortex Code is billed based on consumption. Streamlit in Snowflake runs on compute managed by Snowflake. Practically, that means teams can start building without first having to justify a brand-new software category; spend scales with actual usage.
And finally, there is the model-flexibility point. Frontier models will keep changing. Snowflake’s approach gives teams access to supported models inside the same governed environment. That means you don’t have to redesign the workflow every time the model landscape moves.
The bigger story: from reporting to action
So when people ask me why finance is excited about Cortex Code, my answer is simple: it helps finance operate at the speed the business expects, without asking finance to leave the governed environment where its data already lives.
That is the bigger story here. This is not about making finance more technical for its own sake. It is about making finance more useful. It is about giving analysts the ability to build connected applications, automate repetitive workflows and deliver insight faster. It is about turning finance data into something the business can interact with, not just something it reviews after the fact.
What comes next
Over the next set of posts, my team and I will share specific examples of how our team is using Cortex Code in practice. So bookmark this page and we will try to get the content out as quickly as possible:
Accelerating FinOps: Transforming Reporting into Intelligent Systems with Cortex Code
Automated Variance Analysis for Close Reporting
More to come soon (including long-range planning, automating reporting workflows with outreach, investor relations Q&A)
For finance teams that want to move from static reporting to operational decision support, Cortex Code is not just another AI tool. It is a faster path from data to action.
