Inside a Claude-Built Alternative to a $150K Looker Setup
A three-person data platform team says it built an internal HTML dashboard service around Claude Code and Claude Opus 4.5, storing files in Amazon S3 behind SSO and querying Snowflake with each viewer’s OAuth token. The service reportedly reached as many weekly users as Looker within eight months, while Looker usage fell by one-third. The team also documents sandboxed dashboards, security checks, and its decision to consider Anthropic’s newer Claude Dashboards product if starting from scratch.
The case study offers a concrete example of AI-assisted development replacing parts of an enterprise business-intelligence stack, while showing the substantial security and platform work required.
Like a lot of companies, we pay for Looker. Let’s say $150k a year, the average Looker licence1, roughly 300 seats. We are 3 data platform engineers at a company of about 500 people, and most of our week goes on the unglamorous stuff: keeping Snowflake fed, babysitting a dbt project that passed 2,000 models a while ago, and running the Airflow, Airbyte, DataHub and Kubernetes plumbing underneath it all.
In February, one of us built a thing on the side to host HTML files. It took an afternoon. 8 months later, and it’s already got as many weekly users as Looker in our company, maybe even more and Looker’s down a third.
Tech influencers keep posting about software being dead, and some say they’ve cancelled and vibecoded all their SaaS subscriptions. They say AI is going to kill B2B SaaS.2
SaaS is dead, apparently, so can we replace our enterprise BI tool?
Share of "self-hosted" GitHub repos
Meanwhile, "slop" on Hacker News
Left: share of new GitHub repos described as "self-hosted". Right: share of Hacker News comments saying "slop".3
#Looker#Claude Code#Claude Opus 4.5#Anthropic#Claude Dashboards#Snowflake#Amazon S3#dbt#Apache Airflow#Airbyte#DataHub#Kubernetes