AI &amp; Analytics Insights for Mid-Market Leaders

Mar 19, 2026 · 9 min read

1 Ticket to Change a Price. Not Anymore.

How we moved pricing out of hardcoded macros and into the hands of Finance — no ticket, no PR, no waiting.

Mar 19, 2026 · 12 min read

We Finished 5 Days Early. Here's the System.

How we structured 28 Claude Code sessions across 9 days to deliver a complete FY27 pricing rebuild — 5 days early.

Mar 19, 2026 · 13 min read

CLAUDE.md: How We Made AI Safe on Financial Data

What goes in CLAUDE.md when you're using Claude Code on a financial dbt codebase — and why hooks are a different kind of guardrail.

Mar 20, 2026 · 15 min read

The AI Guardrails We Left Behind

The most useful thing we left behind wasn't documentation — it was a starter repo that let the team use Claude Code safely on day one after we left.

Advisory & Strategy

5x Scope Growth. Zero Scope Creep.

We were hired to migrate 377 legacy BI objects. We ended up building a 161-model analytics platform across 7 domains. Here's why that was the right outcome.

Mar 20, 2026 · 9 min read

Mar 20, 2026 · 18 min read

15 Bugs Hiding in Your Financial BI

Your production financial BI system probably has bugs in it. We found 15 in a well-maintained system — none of them would have been caught by testing alone.

Mar 20, 2026 · 11 min read

Under $2,000 Total Variance. Here's How.

How we achieved under $2,000 total variance — a 0.002% accuracy rate — using parallel dual models, hierarchical grain validation, and strict parity-first discipline.

Mar 20, 2026 · 9 min read

The Permission Tax Nobody Tracks

Snowflake RBAC ate 17–26 engineering hours in the first two months of our engagement. Here are the role patterns — future grants, functional/access split, per-developer isolation — that eliminated most of it.

Mar 20, 2026 · 8 min read

Replicate the Bug. Then Fix It.

Sometimes the correct engineering move is to deliberately write buggy code — and document exactly why.

Mar 20, 2026 · 4 min read

4 Snowflake-Native dbt Constraints the Docs Skip

A practitioner account of running a production dbt project in Snowflake's native runner — including the four constraints the vendor docs skip.

Mar 20, 2026 · 8 min read

The Auth Walls Nobody Documents

How Okta + Snowflake SSO behaves in practice, what externalbrowser means for automation, and what architecture we settled on after hitting every wall.

Mar 20, 2026 · 9 min read

377 Objects. 51 Models. A Framework for the Gap.

How we reduced 377 legacy BI objects to 51 dbt models — and the domain-isolated architecture that scaled them to 161 without collapsing.

Automated Reporting & Dashboards

86 Charts, One Page. Here's How We Fixed It.

How a cloud security company went from 86 charts on a single overloaded page to 6 dashboards across 22 organized pages — and the keep/merge/kill framework that got them there.

Mar 20, 2026 · 14 min read

Mar 20, 2026 · 11 min read

Cortex Analyst on Real Financial Data: An Honest Account.

We ran Snowflake Cortex Analyst against production-validated revenue data built on 51 dbt models. The semantic layer worked — with meaningful caveats worth knowing before you build.

Mar 20, 2026 · 8 min read

Finance Updating Pricing Without Filing a Ticket. Yes, Really.

How we gave a Finance team full pricing update capability — no engineering, no git, no Jira tickets — using Sigma input tables wired to dbt seeds with dropdown validation sourced from live pipeline data.

Advisory & Strategy

Migration as Investment, Not Cost

The real ROI of analytics modernization comes from organizational capability, not just faster dashboards.

Mar 23, 2026 · 5 min read

Mar 23, 2026 · 4 min read

606% ROI: The Numbers Behind BI Migration

How a 5-month analytics migration delivered 606% ROI — with the numbers to prove it.

Mar 23, 2026 · 6 min read

AI in Analytics: Beyond the Hype

What AI actually did — and didn't do — on a 5-month analytics migration with real financial data.