dagctl blog

Technical guides for production data teams

Deep dives on running SQLMesh at scale — infrastructure, CI/CD, scheduling, migration from dbt, and the sharp edges nobody warns you about.

Subscribe via RSS

Latest Engineering

So you want an agentic semantic layer?

A cheap model with a SQL tool found the data at 14, 100 and 1,000 tables. It missed two things: what an ID means, and who is asking.

Scotty Pate Oct 12, 2026 7 min read

Vision

Why You Should Choose SQLMesh Over dbt in 2026

Most of data engineering is three questions: what is the state, will it work in prod, what breaks. SQLMesh answers all three. dbt leaves them to you.

Oct 1, 2026 10 min read

Engineering

Running SQLMesh on Kubernetes: Jobs, CronJobs, and the State Database

Running SQLMesh on Kubernetes: a working CronJob for sqlmesh run, credential setup, connectivity, and the failure modes that show up once it's live.

Sep 18, 2026 10 min read

Engineering

The Parts of SQLMesh That Aren't Production Ready

Run capacity you can't partition, backfills with no guardrail, and prod deploy access that isn't enforced at all. Three things SQLMesh leaves you to build before a large project is safe to operate, and what building them actually involves.

Aug 14, 2026 13 min read

Vision

SQLMesh Is Not Going Anywhere

Fivetran bought Tobiko, then announced a merger with dbt Labs, and everyone asked the same question: is SQLMesh dead? Then SQLMesh went to the Linux Foundation. Here's why that ends the argument — and why the argument was always weaker than it looked.

Jul 31, 2026 4 min read

Engineering

The Model Is Commodity. The System Around It Is Not.

Notes on building autonomous systems around language models. The model is the least interesting part.

Apr 7, 2026 15 min read

Engineering

dbt vs SQLMesh in 2026: Which Tool Is Right for You?

A technical comparison of dbt and SQLMesh, updated for 2026: what the Fivetran and dbt Labs merger and SQLMesh's move to the Linux Foundation change, plus state management, environments, SQL understanding, change workflows, and when each tool fits.

Mar 25, 2026 18 min read

Vision

Data Engineering, Industrialized

No car company starts by smelting its own steel. The competition wouldn't allow it. But every data engineering team starts from zero. That's the gap, and it's closing.

Mar 18, 2026 6 min read

Engineering

SQLMesh for dbt Users: The Mental Model Shifts That Matter

A practical guide for experienced dbt users evaluating or migrating to SQLMesh — covering the real differences in how environments, deployments, state, and incremental models work.

Feb 27, 2026 11 min read

Engineering

Operating SQLMesh in Production: State, Scheduling, and Observability

How SQLMesh state works under the hood, how to set up scheduling with the built-in scheduler or external orchestration, and what to monitor in a running production system.

Feb 20, 2026 12 min read

Engineering

Running SQLMesh in Production: Infrastructure, CI/CD, and Deployment

How to containerize SQLMesh, set up CI/CD pipelines with the plan workflow, configure multi-environment deployments, and avoid the infrastructure mistakes that bite teams in production.

Feb 13, 2026 14 min read