How a Waste Firm Rebuilt Its 10-Year Plan (and Made AI Listen)

Long-term planning sounds straightforward until you're the one doing it. For a hazardous waste disposal company, it's not about next quarter's revenue. It's about 10 years out, across 40-plus legal entities, each with over 100 cost centers and hundreds of spending categories. That granularity isn't a luxury. It's the only way the forecast can serve the tax team, treasury, HR, and the C-suite without falling apart.

Tax wants the split between goods and services, the legal entities, the jurisdictions. Treasury needs a cash view. HR cares about headcount assumptions. Executives want to see the trade-offs between growth, margins, investment, and free cash flow. One model has to feed all of them.

So, like most companies, when the business outgrew its tools, the finance team did what finance teams do: they built a massive Excel workbook. It worked, but it turned into a Frankenstein. New tabs piled up. Formulas got patched. Logic was layered on top of old logic. The model became more valuable and harder to maintain at the same time.

That was the problem they set out to solve.

The Spreadsheet Monster Gives Way

A little over a year ago, the team rebuilt the long-term forecast on Snowflake, with Streamlit as the interface. They called it Snowplan. The goal wasn't to build a dashboard. It was to build a planning platform that felt intuitive to finance users but had the compute power, governance, and scalability of Snowflake underneath.

Now, analysts update assumptions directly in the app. Those changes write back to Snowflake, the model runs, and the outputs appear instantly. No more broken formulas. No more saving files back and forth. No more guessing which version is the source of truth.

That shift changed the whole planning process. Instead of maintaining a giant offline workbook, they now have an application connected to the actual source data. Actuals flow in automatically. Assumptions are versioned. Scenarios can be compared. Different roles see the same platform at different levels of detail.

Individual contributors get fine-grained input pages and scenario management. Managers get visibility into changes and approvals. Executives get a consolidated P&L, free cash flow, and key scenarios. Long-term planning isn't just a modeling exercise—it's an alignment process. The less time finance spends maintaining the model, the more time they spend actually working with the business on strategy.

Putting the Model Where the Data Lives

The most important decision was putting the model on Snowflake, where the data already lived. That means the model is naturally connected to the original data sources and governed data models. No more manually updating actuals. No more reconciling offline pulls. The model sits right inside the environment that already holds the financial data, permissions, logic, and history.

That brings a few clear advantages.

• It scales. A 10-year forecast spanning entities, cost centers, spending categories, headcount, revenue, balance sheet, and free cash flow generates a lot of data. That's exactly what Snowflake handles best.
• It's governable. Role-based permissions and row-level security mean different users see only what they should. Executives don't need the same interface as analysts, and analysts don't need to export separate versions for every stakeholder.
• It's reusable. The same platform can support headcount planning, equity modeling, treasury cash forecasting, hedging, legal entity forecasts, COGS planning, and M&A scenarios.

CoCo Makes Scenario Planning a Conversation

Streamlit made Snowplan usable and scalable. Snowflake CoCo made it conversational. Before CoCo, users had to know where to click, which assumption to change, and how to understand the downstream impact. CoCo changed that.

Now, instead of hunting through pages of assumptions, you can just ask a question in plain English. Ask CoCo to compare two versions of the forecast and summarize the key drivers. Ask what changed between the plan you showed the board last year and the latest version. Ask about the net impact, the drivers of margin expansion or dilution, and which assumptions deserve the most attention.

In executive planning, that's powerful. The real question isn't "Can you give me the latest numbers?" It's "What changed, why, and what does that mean for our narrative?" CoCo turns that analysis into a conversation—and lets finance iterate while the strategic discussion is still happening.

A Real Example: Scenario Planning for a Potential Tax Change

One of the best examples is scenario planning around a potential tax change. In the old world, this would start with a meeting. The tax team defines the problem, identifies the affected sales, pulls the data, builds assumptions, updates the model, reviews the output, and runs sensitivity tables. Then they figure out who else needs to be involved.

With CoCo in Snowplan, it's much smoother. You start by asking CoCo to summarize the potential tax change. Then you ask it to create a new forecast version assuming the change goes through. That immediately leads to the kind of back-and-forth a finance team would have in a meeting: Is the tax passed on to customers, or absorbed as a margin hit? What percentage can realistically be passed through? Which sales are affected? What's the impact on revenue, gross margin, operating margin, and free cash flow?

Because the analysis sits on Snowflake tables, CoCo can identify which sales would be impacted, output the financial effect, and show the core metrics behind the numbers. It can also create sensitivity tables that show how operating margin dilutes under different pass-through assumptions.

Equally important, it flags risks and caveats. For example, a first-order model might not capture the extra indirect costs of supporting filings, maintaining compliance data, or meeting new reporting requirements. That's the kind of nudge a good finance partner gives before anyone treats a scenario as a conclusion.

CoCo can even help generate next steps, like drafting an email to the tax team summarizing the analysis, key assumptions, open questions, and decision points. The system isn't just spitting out a number. It's helping frame the problem, identify the right experts, and push the process forward.

Why This Matters for Finance Teams

Finance teams are constantly asked to answer strategic questions faster than traditional planning cycles allow. What if we accelerate growth? What if we open a new office? What if cloud costs improve by 25 basis points? What if compensation inflation runs higher than expected? What if a tax or regulatory change hits part of our sales? What if we reallocate investment across functions?

These aren't hypotheticals. They're real questions that executives throw out in real time. The problem is that traditional planning tools and giant spreadsheets aren't built for that kind of iteration. They're built to produce one plan, not to support an ongoing strategic dialogue.

By building Snowplan on Snowflake with Streamlit, this team created a platform that scales with business complexity. Adding CoCo made it conversational. That combination changes what finance can actually do. Instead of spending hours updating actuals, maintaining formulas, reconciling scenarios, or comparing versions by hand, they spend time challenging assumptions, aligning executives, evaluating trade-offs, and shaping long-term strategy.

Trust Is the Foundation of AI-Driven Planning

For conversational planning to work in finance, the numbers have to be trustworthy. That's why the architecture matters. CoCo isn't generating forecasts out of thin air. It's interacting with the same governed data, assumptions, and logic that power Snowplan. When it compares versions or explains drivers, it's relying on the same Snowflake data models and planning logic the team already uses.

Every scenario is versioned. Every change can be reviewed. Access control follows the app's role model. Analysts and executives can see what changed, understand why, and roll forward or roll back if needed. This is a key distinction. They're not asking leaders to trust a black box. They're using AI to operate a governed planning platform where data, logic, permissions, and outputs are visible, explainable, and auditable.

From Planning Tool to Strategic Platform

The most exciting part is that Snowplan has grown beyond its original use case. Once the model was on Snowflake, the architecture became reusable. The same foundation now supports headcount planning, equity incentive modeling, treasury cash forecasting, hedging, legal entity forecasts, COGS planning, and M&A scenario analysis.

That's the advantage of building a platform rather than a one-off application. Each new planning workflow reuses the same governance backbone, connects to the relevant data sources, and exposes a finance-friendly Streamlit interface. And with CoCo, each process is easier to query, adjust, and explain in natural language.

More finance teams will likely follow a similar path: first, move the model to where the data lives. Second, build an intuitive app layer for users. Third, use AI to make the planning process conversational.

The Real ROI

The real ROI of Snowplan isn't making the finance team more technical. It's giving them back time for judgment. Long-term planning isn't about maintaining a spreadsheet. It's about understanding where the business is headed, helping leaders decide where to invest, how to balance growth and profitability, which risks are emerging, and which trade-offs matter most.

Snowflake and Streamlit gave the team a platform that scales, stays governed, and connects to real-time data. CoCo is making it faster, more interactive, and more strategically valuable. The result: finance spends less time updating models and tweaking assumptions, and more time with the executive team iterating on the company's long-term strategy.

For FP&A teams, that's the real value of AI in planning. It's not replacing finance. It's taking away the manual labor that slows them down, so they can focus on what they're actually supposed to do.