Tag1 Consulting

When Good Links Go Bad: How AI Cut Link Verification in Drupal’s Metatag Module from Hours to Minutes “Sammy Gituko shows how AI sped up fixing broken links in Drupal’s Metatag module, improving documentation reliability and contributor efficiency.”

We just cut a 3-hour task to 15 minutes. The AI couldn't make the contextual decisions and needed human judgment for the parts that mattered most. Learn more here: http://tag1.co/wro

#ArtificialIntelligence #AI #Drupal #SoftwareDevelopment #DigitalTransformation #WebDevelopment

Exploring the Future of the Drupal Group Module Tag1's Kristiaan Van den Eynde, creator of the Drupal Group module, previews Group 4: ecosystem consolidation, redesigned workflows and public/private Groups.

Dive into the future of #Drupal's Group Module with Michael Meyers and Kristiaan Van den Eynde on our latest Tag1 TeamTalk! Discover insights on module evolution, architecture, and the art of managing large-scale software. A must-listen for anyone in the Drupal community. http://tag1.co/77n

Exploring the Future of the Drupal Group Module Tag1's Kristiaan Van den Eynde, creator of the Drupal Group module, previews Group 4: ecosystem consolidation, redesigned workflows and public/private Groups.

Explore the evolution and future of Drupal's #Group module with insights from maintainer Kristiaan Van den Eynde. Join us for a revealing conversation on Tag1 Team Talks. http://tag1.co/77n
#Drupal

The Practical Path to AI Search A practical guide to AI Search: Google‑style conversational UX and sentiment intelligence on your stack, no vector database required.

The Practical Path to AI Search. A clear, practical look at how modern site search can deliver better answers, better experiences, and better insight without added complexity. Built from real production work at Tag1. Read the Guide: http://tag1.co/h3j
#Scalability #UserExperience #WebDevelopment #AI

Mastering the Data Organization Matrix (DOM): The Insights Supply Chain Framework - Part 3 Tag1’s Dr. Duru Ahanotu explains the Data Org Matrix, data as a core competency, and how to structure modern data‑driven organizations.

Check out the latest episode of #Tag1TeamTalks as we learn more about the Insights Supply Chain Framework with Dr. Duru Ahanotu. Discover the Data Org Matrix and its impact on data-driven decisions in your organization. http://tag1.co/1vh

The Practical Path to AI Search A practical guide to AI Search: Google‑style conversational UX and sentiment intelligence on your stack, no vector database required.

People expect search to explain, not just list pages. This guide shows how to build faster, clearer search experiences that help users find answers, and help teams understand what's working.
The Practical Path to AI Search: http://tag1.co/h3j
#Scalability #UserExperience #WebDevelopment #AI

Centralized vs. Decentralized Data Teams: The Insights Supply Chain Framework Part 2 Part 2 of Michael Meyers' and Dr. Duru Ahanotu's talk about the Insight Supply Chain, focusing on how orgs choose centralized vs. decentralized data teams

Want to leverage your data for better insights? Check out our Tag1 TeamTalk podcast with Dr. Duru Ahanotu, who discusses the Insight Supply Chain Framework and the impact of team organization on your data strategy. http://tag1.co/sct

#DataStrategy #BusinessStrategy

Speed up your testing with Deuteros Learn how AI helped, Francesco Plcella, speed up Drupal testing with Deuteros, making open‑source development faster and more accessible.

Tag1's Francesco Placella shows how AI helped him build Deuteros, an open‑source PHP library that lets you unit‑test Drupal entities without dragging kernel tests into every... https://www.tag1.com/blog/speed-up-testing-with-deuteros/
#AIApplied #Drupal #PHP #Testing #DeveloperExperience #OpenSource

Speed up your testing with Deuteros Learn how AI helped, Francesco Plcella, speed up Drupal testing with Deuteros, making open‑source development faster and more accessible.

We love performance! Tag1's Francesco Placella shows how AI helped him build Deuteros, an open‑source PHP library that lets you unit‑test Drupal entities without dragging k... https://www.tag1.com/blog/speed-up-testing-with-deuteros/

#AIApplied #Drupal #PHP #Testing #DeveloperExperience #OpenSource

The Three S’s: How I Think About AI Agent Security Learn how Fabian Franz’s 3S model helps you protect your data from overpowered AI agents by limiting calendar, email, and tool access.

Hidden instructions are the new phishing links. Fabian Franz shows how a simple restaurant review or calendar invite can jailbreak an agent that has access to your email...
https://www.tag1.com/blog/how-to-think-about-ai-agent-security/

#ArtificialIntelligence #AISecurity #LLMSecurity #Agents #Tag1

Part 3: Model Architecture, Building an LLM from Scratch in Rust Learn how GPT-2 works under the hood by implementing its full forward pass and transformer blocks from scratch in Rust with Feste.

RNNs wait in line. Transformers run in parallel.
Part 3 shows how attention + causal masking let GPT‑2 process 1,024 tokens at once while keeping long‑range context—implemented...
https://www.tag1.com/how-to/part3-model-architecture-building-an-llm-from-scratch/

#AIEngineering #RustLang #OpenSource

The Three S’s: How I Think About AI Agent Security Learn how Fabian Franz’s 3S model helps you protect your data from overpowered AI agents by limiting calendar, email, and tool access.

One AI agent that reads your calendar, browses the web, and acts "as you" is incredibly convenient, and a perfect attack surface.
In Tag1's lFabian Franz explains AI agent security with the three S's. https://www.tag1.com/blog/how-to-think-about-ai-agent-security/

Part 3: Model Architecture, Building an LLM from Scratch in Rust Learn how GPT-2 works under the hood by implementing its full forward pass and transformer blocks from scratch in Rust with Feste.

Part 3 of Tag1's LLM series builds a full GPT‑2‑style transformer: embeddings, multi‑head attention, layer norm, residuals, GELU MLPs, 12 blocks, 87M params.

https://www.tag1.com/how-to/part3-model-architecture-building-an-llm-from-scratch/

#AI #LLM #Rust #Transformers #MachineLearning

Using AI to Upgrade a Simple Frontend VAT Tool Janez uses AI to refactor a VAT web app, upgrading a vanilla JS Excel‑to‑XML tool with TypeScript, Vue, and experiments in Elm.

AI isn't magic—it's leverage.

In this AI Applied post, we used AI to refactor a monolithic JS app, migrate it to VueJS, improve styling, and test new approaches fast.

It worked.

It failed.

And the balance mattered.

Read the full post: tag1.co/nun

Beyond Pretty: The Real Work of UX/UI Discover how Tag1’s UX/UI leaders use research, collaboration, and iteration to turn subtle UX friction into measurable business results.

From research to iteration, discover how Tag1's UX/UI leaders use research, collaboration, and iteration to turn subtle UX friction into measurable business results in our latest #Tag1TeamTalk! Tune in today! https://www.tag1.com/tag1-team-talks/beyond-pretty/

Navigating Local Development Landscapes: Tools, Benefits and Best Practices Learn how DDEV simplifies local development, speeds onboarding, reduces debugging pain, and helps teams reliably match production environments.

Is your web development process living in the past? Tune in to #Tag1TeamTalks to explore modern local development environments with industry experts and see why tools like #DDEV are essential for today's developers.
https://www.tag1.com/tag1-team-talks/navigating-local-development-landscapes/

Building Tactician: PHP Scheduling with AI AI-assisted PHP scheduling library for esports and tournaments, built for Mission Gaming using Cline. Learn how Jamie Hollern used AI to create Tactician.

Combinatorial scheduling gets wild fast.

This post shows how AI helped design fair, deterministic tournament scheduling in PHP, without solver engines.

Worth the read: https://www.tag1.com/blog/building-tactician-with-ai/

#AIApplied #AIInPractice #PHPDev #EngineeringLessons

Part 2: Tensor Operations, Building an LLM from Scratch in Rust Explore how tensor math powers transformers in Rust. Learn how Feste uses tensors to build the foundation of a GPT-2 style language model.

Transformers aren’t magic; they’re tensor math.
In Part 2 of Feste, we build a tensor library from scratch in Rust and show how attention, normalization, and embeddings emerge from real memory layouts and CPU constraints.
#AI #AIApplied #Transformers #RustLang #Systems
www.tag1.com/how-to/part2...

How to deal with AI Agents that Phone It In Learn how AI workflow orchestrators prevent planner fatigue in multi-agent LLM systems, ensuring reliable, repeatable code audits and research tasks.

Ever feel like your tools aren't giving 100%? Scott Hadfield shows how Tag1 keeps multi-step AI work reliable and accurate.
https://www.tag1.com/blog/how-to-deal-with-ai-agents/

#AI #AIApplied #Workflow #DigitalTools

Building Tactician: PHP Scheduling with AI AI-assisted PHP scheduling library for esports and tournaments, built for Mission Gaming using Cline. Learn how Jamie Hollern used AI to create Tactician.

AI didn't just speed this up, it made it possible. Read how it worked : https://www.tag1.com/blog/building-tactician-with-ai/

#AIApplied #PHP #SoftwareEngineering #OpenSource #DevStories

Turning Data into Decisions: The Insights Supply Chain Framework Discover the Insight Supply Chain Framework to optimize your data, AI, and analytics, turning insights into actionable business value.

From Yahoo to Tag1, Dr. Duru Ahanotu's journey to creating the Insights Supply Chain Framework offers a compelling story and impactful solutions. Listen to our podcast and learn how to l... https://www.tag1.com/tag1-team-talks/turning-data-into-decisions/

#DataStrategy #PodcastSeries #Tag1TeamTalks

Part 1: Tokenization, Building an LLM From Scratch in Rust Learn how to build a language model from scratch in Rust, starting with part 1 of 6: tokenization, BPE, and vocabulary trade-offs.

If you want to understand how a language model truly works read part 1 of our 6 part series.
https://www.tag1.com/white-paper/part1-tokenization-building-an-llm-from-scratch-in-rust/

#OpenWeb #TechCommunity #MachineLearning #DeepLearning #LLM #AIApplied #AI #Rust

How to deal with AI Agents that Phone It In Learn how AI workflow orchestrators prevent planner fatigue in multi-agent LLM systems, ensuring reliable, repeatable code audits and research tasks.

AI can do amazing things, but sometimes it slacks off. See how workflow orchestrators keep tasks on track and make sure nothing gets skipped. https://www.tag1.com/blog/how-to-deal-with-ai-agents/

#AI #AIApplied #TechTips #Productivity

Building QR Reader Light: My First Android App Ra Mänd built QR Reader Light, a minimalist Android QR scanner, using AI to develop quickly in Kotlin, a language he had never used before.

Ever wondered what it's like to build your first Android app with the help of AI? Tag1 developer Ra Mänd used Claude Sonnet 4 and Cline, he created QR Reader Light, no-ads QR scanner in less than 20 hours.

https://www.tag1.com/blog/building-qr-reader-light-with-ai/

Turning Data into Decisions: The Insights Supply Chain Framework Discover the Insight Supply Chain Framework to optimize your data, AI, and analytics, turning insights into actionable business value.

Join Michael Meyers and Dr. Duru Ahanotu as they explore the ground-breaking Insights Supply Chain Framework. Understand why it's time to rethink traditional Business Intelligence. Catch the new episode now!
https://www.tag1.com/tag1-team-talks/turning-data-into-decisions/

#DataScience #Podcast

But to answer your question specifically - "Rust gives us low-level control over memory and performance, and the borrow checker forces us to think clearly about ownership and memory safety. "

Great question! We are great fans of @rust-lang.org and have been using it for years -- in particular for our Goose load testing framework, which you can learn more about at Tag1.com .

Part 1: Tokenization, Building an LLM From Scratch in Rust Learn how to build a language model from scratch in Rust, starting with part 1 of 6: tokenization, BPE, and vocabulary trade-offs.

Part 1 of our 6 part series on building a language model is now live. Read Part 1: https://www.tag1.com/white-paper/part1-tokenization-building-an-llm-from-scratch-in-rust/

#TechCommunity #MachineLearning #LanguageModels #DeepLearning #OpenSource

Building QR Reader Light: My First Android App Ra Mänd built QR Reader Light, a minimalist Android QR scanner, using AI to develop quickly in Kotlin, a language he had never used before.

We just published a new post in Tag1's AI Applied series:
Developer Ra Mänd used Claude Sonnet 4 + Cline to build QR Reader Light, a clean, ad-free QR scanner for Android.

What's cool about it:
– B... https://www.tag1.com/blog/building-qr-reader-light-with-ai/

#AI #Android #Kotlin #Tag1 #AIApplied

Enhancing Goose Load Testing with an AI-Powered Killswitch AI-powered killswitch improves Goose load testing safety and control. Learn how Tag1 used Claude to build safer automated prewarming and performance workflows.

Automated load testing is safer with a killswitch.
Claude Code assisted in building and refining Goose's new control feature.

Learn more: https://www.tag1.com/blog/enhancing-goose-load-testing-with-ai/

#Goose #LoadTesting #WebPerformance #DevOps