SiriusOne:
Modernizing BI & Data Engineering with Claude Code
How RG Brands and SiriusOne built Claude Code into Qlik and ClickHouse engineering work, cutting core BI & data engineering delivery from five engineers to two and average troubleshooting time from four hours to 30 minutes.Client & Project Overview
RG Brands is modernizing its BI and analytics environment. The program covers Qlik development and reporting, moving analytics workloads to a DLH / ClickHouse target stack, consolidating reports and improving infrastructure.
SiriusOne is helping by redesigning selected BI and data engineering workflows around Claude Code.
Claude Code is part of the team’s daily engineering work rather than a separate experiment. Engineers use it for SQL development, Qlik scripting, debugging, refactoring, technical documentation and review, and they remain responsible for validation and production changes.
Business Problem
Modernizing an analytics platform while still shipping production reports is a hard balance. The team has to migrate workloads, keep existing dashboards running, fix issues, document changes and pass knowledge on, all without letting delivery slip.
RG Brands needed more than an AI assistant for its engineers. It needed AI-assisted engineering that was repeatable, reviewable and safe enough for everyday delivery.
So SiriusOne built Claude Code into the engineering lifecycle itself. The aim was to cut repetitive work, make development and documentation more consistent, and help the BI and data teams move faster while engineers keep technical oversight.
Tech Stack
AI-assisted engineering: Claude Code
BI & Analytics: Qlik
Data Engineering: SQL · DLH · ClickHouse
#ClaudeCode
#AISDLC
#BusinessIntelligence
#Qlik
#ClickHouse
Project Timeline
Duration
~6 months
Effort
~2,000 hours
Foundation
Month 1
Assess selected workflows, establish operating practices and introduce Claude Code-enabled engineering patterns.
Migration
Months 3–4
Apply and refine approved workflows across DLH / ClickHouse migration, Qlik development, troubleshooting and documentation.
Scale
Months 5–6
Standardize the practices that work, expand adoption across the team and strengthen documentation and technical handover.
Team involved
ETL/BI Developer
SQL development, DLH / ClickHouse migration, query optimization and troubleshooting.
BI Developer/Analyst
Qlik scripting, data-model analysis, reporting and technical documentation.
Solution Overview
How We Worked: AI-Assisted Engineering, with Human Control Built In. Claude Code is used in selected engineering workflows, each with defined review and validation steps. Engineers use it to speed up analysis, implementation and documentation. They stay accountable for architectural decisions, production changes and final quality.
Claude Code-enabled engineering workflows
Claude Code supports selected Qlik scripting, SQL development, data-model analysis, debugging and refactoring. Engineers also use it to work through existing SQL, scripts and technical context, narrow down likely causes of issues and draft fixes. SiriusOne turns this into repeatable workflows instead of ad-hoc prompting, and engineers validate the result.
Knowledge captured as part of delivery
The team writes technical decisions, documentation and handover materials as part of the engineering work. Less depends on knowledge that was never written down, and practices that work are easier to reuse.
Repeatable engineering practices
When a way of using Claude Code works, it becomes a shared practice for development, troubleshooting, review and documentation instead of staying one person’s prompting habit.
Human-reviewed production changes
AI-generated code and recommendations never go straight to production. Any change that affects production goes through engineering review, testing and validation.
Results
5 → 2 engineers
Core BI & data engineering delivery went from five engineers to two.
4 h → 30 min
Average troubleshooting time for selected Qlik, SQL and data engineering issues, down from about four hours.
40% faster
Execution of the top-10 optimized queries.
30% lower
Processing cost for optimized queries.
Similar
implemented cases:
Modernizing BI & Data Engineering with Claude Code
Scalable Data Lake for Enterprise Analytics
Automated ETL Pipeline for Financial Data Processing
Real-Time Data Processing for E-Commerce Insights
