As organizations scale their data platforms on Snowflake, they unlock new levels of flexibility, performance, and AI-driven innovation. At the same time, understanding what drives consumption and where costs originate becomes critical—not only for optimization but for informed business decisions. With the help of Snowflake CoCo, Sigma Software empowers clients with deeper insights into cloud spending, enabling a new level of cost transparency and operational efficiency.
Use Case 1: Restoring Cost Transparency in a Multi-Layered Architecture
Use Case 2: Turning Cost Monitoring into a Self-Service Capability
As a Snowflake AI Data Cloud Services Partner, Sigma Software works with companies that rely on data and AI in core business processes. We help them build systems that are secure, scalable, and designed for long‑term operation. And we take every opportunity to streamline developing Snowflake‑based solutions and facilitate their operation for our clients.
One of useful tools for this is Snowflake CoCo, an AI agent purpose-built for the Snowflake data stack.
What is Snowflake CoCo
Snowflake CoCo represents a new generation of tools within the Snowflake ecosystem.
Unlike traditional approaches that rely heavily on manual SQL and fragmented tooling, CoCo operates as a native AI assistant embedded directly into the Snowflake Data Cloud.
For Sigma Software teams, CoCo capabilities are already driving tangible productivity gains. Our data engineers have leveraged it to enhance cloud cost transparency, enabling clients to clearly understand how their resources are utilized and where their budgets are allocated.
Below, we explore two practical Snowflake CoCo use cases that highlight its potential in optimizing and managing Snowflake costs.
Use Case 1: Restoring Cost Transparency in a Multi-Layered Architecture
The challenge: “We see the bill—but not the story behind it”
In modern setups, data costs are driven by multiple factors: compute (queries, pipelines), storage, background processes.
The situation becomes even more complicated when optimization tools sit between users and the platform, automatically routing workloads and reshaping execution flows. It results in cost aggregation, which obscures execution paths. Thus, attributing costs can become really difficult.
Boris Trofimov, a seasoned Data Architect with over 20 years of experience in the field faced this exact problem. He asked it to break costs down to original workloads, identify their true source, and create a usable cost model.
“I described the problem and fed it into CoCo—and it generated a fully working solution, including SQL that accounted for all the nuances of our setup,” says Boris.
“It allowed us to reconstruct the original cost distribution across warehouses by analyzing query history and thus restore transparency in Snowflake billing. The solution we built wasn’t just a one-time query—we shared it with the team so others could reuse it and automate their cost tracking.”
The impact: From hours to minutes
What would normally require deep investigation and manual query building was completed in a fraction of the time.
- ~15–20 minutes total effort
- Working, production-ready logic
- Reusable across teams
The generated solution was shared with analysts and finance stakeholders, becoming part of broader cost visibility practices. The key shift here is not automation—it’s augmented problem-solving, where Snowflake CoCo helps build the solution, not just assist with tasks.
Use Case 2: Turning Cost Monitoring into a Self-Service Capability
The challenge: Continuous cost awareness
In Snowflake environments, getting a complete picture of what is happening often requires data engineers to pull information from multiple system tables and manually correlate it with query activity. Moreover, supporting custom analytics typically means rewriting or adjusting queries each time new insights are needed.
This process is time-consuming and typically requires deep expertise, making it difficult to scale across teams. Snowflake CoCo significantly streamlines this effort by enabling faster, more flexible custom analytics. It can generate reports based on resource usage, identify warehouse optimization opportunities, and correlate usage patterns with client traffic—delivering deeper, more actionable insights.
Our Data Science Engineer Vlad Chaikovsky, who uses Snowflake CoCo on a daily basis, notes:
“It correlates inputs and produces actionable insights much faster, freeing up engineers’ time for system optimization.”
The team decided to go further. They used CoCo to build a custom AI agent for cost tracking based on reusable logic and skills. Instead of relying on ad hoc analysis, they created a structured way to interact with cost data. And they shared the agent to finance and business roles giving them a possibility to simply ask the questions they need—and get answers instantly.
The impact: Making cost insights accessible and actionable
With this approach, cost monitoring moves from a technical task to a shared capability across the organization.
The agent tracks billing and credit consumption, answers cost-related questions on demand, provides structured insights in a predefined format.
The benefits it provides:
- Broader access to cost data beyond engineering teams
- More consistent and comparable reporting outputs
- Faster, more intuitive interaction with cost insights
The key shift is accessibility: cost visibility becomes embedded into everyday workflows, enabling more informed decisions across both technical and business teams.
Final Thought: Where Technology Meets Expertise
The true value of tools like Snowflake CoCo does not lie in automation alone—it emerges at the intersection of advanced AI capabilities, domain expertise, and a structured engineering approach.
As demonstrated by Sigma Software teams, approaching AI not as a shortcut but as an integral part of the development workflow enables faster delivery, more effective problem-solving, and the ability to tackle increasingly complex challenges. Engineers remain in control—defining problems, validating results, and shaping solutions—while AI acts as a powerful multiplier, enhancing efficiency and enabling new ways of working with data.
By combining the right tools with the right expertise, Sigma Software is not only optimizing existing processes but also unlocking new capabilities for our clients—from real-time cost visibility to scalable, self-service data practices that empower both technical and business stakeholders.
