erickohlenberg.com

    • A Compute Commodity Clearinghouse

      The AI compute market is, by most measures, a mess. GPU rental prices fluctuate by an order of magnitude across providers. Procurement is opaque — often involving phone calls, NDAs, and negotiated contracts rather than transparent price discovery. Large clusters are rarely available on-demand, and buyers have no standardized way to compare an H100 against…

    • Designing Your Data Model with Jumpstart

      Everything Jumpstart generates flows from a single CSV file. The quality and completeness of what you get out depends entirely on how well you describe your data model going in. This post is a practical reference for writing that metadata file — covering the CSV format, the three foreign key types, view synthesis, and the…

    • Inside the Jumpstart Generator

      Code generators are easy to dismiss as glorified string concatenation — some if statements and a few for loops stitched together with printf. Jumpstart is not that. Understanding how its generator actually works reveals a thoughtfully engineered pipeline with a proper in-memory model, a real templating engine, and a graph algorithm borrowed from compiler theory….

    • Jumpstart and AI: A Foundation for the Modern Development Stack

      AI coding tools have changed the economics of software development. Tools like Claude Code, GitHub Copilot, and ChatGPT can generate working code from a natural language description, explain a complex codebase in plain English, and implement a feature in minutes that might have taken hours to write by hand. The question for development teams is…

    • Start with a Minimum Viable Solution

      Start with a Minimum Viable Solution

      Borrowing from Eric Reis’s book Lean Startup, start your data warehousing project simply.  Often, we envision a comprehensive, elaborate solution, but it goes beyond the problem at hand.  Or, the problem at hand is complex but you may not see a clear path to the solution.  Start small by identifying the smallest possible deliverable that…

    • Alternative Data Warehouse Key Design

      Alternative Data Warehouse Key Design

      I was once on a project where the users waited on the data warehouse load to complete before going home. The original requirement was that the load take no more than 20 minutes. We were never faster than 90 minutes. After a year trying to get the process to go faster, we made the decision…

    Follow Me On Instagram

    @WillamSmith

    This is how it all started…

    Lorem ipsum dolor sit amet, consectetur adipiscing elit. Maecenas sed fringilla velit, eget pulvinar lacus. Nulla risus nunc, accumsan eget eros eu, finibus efficitur leo. Fusce eget vehicula est, ac auctor augue. Praesent tincidunt non nulla eu aliquet. Mauris libero nisl, pellentesque et consectetur vitae, pulvinar eget massa. Quisque non pharetra ex. Nam quis ipsum luctus, consectetur elit nec, interdum justo. Vivamus ac cursus purus. Pellentesque in justo mauris. Vivamus vitae imperdiet nisl. Ut eget leo sollicitudin, rutrum est id, sagittis turpis.

    Quisque iaculis rhoncus sem et elementum. Nullam non ante consequat, aliquet sapien at, tincidunt elit. Aliquam pharetra lobortis lorem, ac scelerisque metus pretium quis. Maecenas vitae dui vel tellus viverra tincidunt in et quam. Aliquam sed urna id lacus tempor mollis vel nec nisi. Phasellus sed urna nec ligula egestas accumsan sed eu ante. Mauris sit amet nisi lorem. Curabitur feugiat ante sed eros tincidunt, ac facilisis enim iaculis. Nulla sodales ac lectus sit amet dictum. In volutpat urna vel leo egestas sagittis. Praesent ornare vulputate massa, non bibendum libero egestas in. Suspendisse semper, ante et faucibus aliquam, risus ligula bibendum eros, id varius lacus eros nec leo. Sed sit amet nibh in lorem ornare congue. Mauris nec massa auctor, interdum erat vitae, tristique tortor. Aliquam accumsan nec velit ut volutpat.

    William Smith

    Writer & Journalist