Notes from the build

Frontier models are expanding what clients can do themselves and what experts can deliver. Good consulting now requires an honest, current view of where each side adds value.

Cheap, capable models make large-scale data enrichment practical. Here is how we turned more than 10,000 incomplete records into structured training examples in minutes.

The biggest AI coding gains are not coming from better prompts. They are coming from custom loops that turn large migrations into bounded, reviewable, and repeatable systems.

We partnered with a software engineering agency to build a full data stack for a leading San Francisco venture capital firm, breaking their dependency on a black-box enrichment vendor and giving them a unified view of people, companies, and signals for the first time.

Mar 19, 2026·Marco Porracin Dieguez

Featured

We helped Takenos stand up its modern data architecture, align product and business metrics, and enable the team to own dbt after handoff.

How a mission to track every commercial flight in Argentina in high-frequency helped bridge the gap between software engineering and the world of data modeling.

Jan 16, 2026·Marco Porracin Dieguez

Featured

How we helped Kiwi migrate to Snowflake, standardize their data processes, and empower their Data/ML teams with a scalable, secure foundation.

We built a Claude plugins marketplace template in a few hours. Here's why your team should start creating and sharing specialized AI agents, and how to get started.

At Blueprint, we believe in open source at the core. Explore our latest contributions to the Singer ecosystem, including taps for Dune Analytics, Firestore, Persona, and Turso.

Why would a data consulting company build a calendar sync tool? We dive into the story behind gcalsync.io and what we learned by building a micro-SaaS from scratch.

Explore how Apache Airflow acts as the central nervous system of our data stack, orchestrating complex workflows, managing dependencies, and ensuring reliability.

Learn how dbt revolutionizes the data transformation process by bringing software engineering best practices like testing, documentation, and version control to analytics.

Dive into our ELT process with Meltano, a powerful open-source tool that helps us manage data extraction and loading from various sources efficiently.

Learn why Infrastructure as Code (IaC) is crucial for a modern data stack and how we use Terraform to manage our platform consistently and reliably.

An introduction to our blog series on building a production-grade data platform from the ground up, component by component.

A practical comparison of Meltano and Fivetran for teams that care about ownership, custom sources, and open ELT. When managed connectors win, and when open ingestion is the better foundation.

Guía de consultoras de analytics engineering y modern data stack en Latinoamérica y Argentina: criterios, shortlist y cuándo conviene un partner con ownership y dbt.

Embedded data and growth squads versus a build-and-handoff analytics engineering engagement. A decision guide for teams choosing how they want to buy data help.

A clear definition of analytics engineering, how it differs from classic BI consulting and data engineering, and when growing teams need governed models and shared metrics.

Why live Supermetrics and Looker Studio connectors break at scale, and how a BigQuery marketing warehouse with shared metric logic fixes speed, consistency, and AI readiness.

A practical shortlist of analytics engineering and modern data stack consultancies in Latin America and Argentina, with criteria for owned platforms, dbt depth, and clean handoff.