Applied AI
Claude Academy: AI Fluency for Builders
A short note on finishing Claude Academy and why practical AI training matters more than abstract noise.
August 26, 2026 · Applied AI
Practical notes on AI agents, dashboards, forecasting, pricing, and the systems behind better decisions.
A short note on finishing Claude Academy and why practical AI training matters more than abstract noise.
A practical note on using conversion rate as a starting point in B2B planning, then layering context before turning pipeline into a forecast.
A practical note on aligning sales, finance, and marketing when everyone wants growth but incentives are different.
A practical note on dashboards, slides, and analysis: start from the decision, not from the pile of data.
Most people don't need five agents. They need one agent that actually gets used. A case for starting simple with AI agents, and why a single working agent beats a team of unused ones.
A simple guide to local AI models, what runs on everyday computers, when a dedicated GPU helps, and why most people should start with what they already own.
How to use AI safely with sensitive data: local models, hardware tiers from laptops to DGX Spark, and data redaction techniques that eliminate exposure risk while delivering real results.
At Dell, a quarterly planning process that once took 2 people working 80+ hours a week is now run by 1 person in a regular 8 to 5. Same 90%+ accuracy. 75% fewer human hours. Billions of dollars in impact.
How Felipe Postigo Gonzalez got started with AI agents using Hermes by Nous Research. No coding required, step-by-step guide, real examples of what you can build.
I built an AI agent that manages my grocery list using Android ADB, a local LLM, and purchase frequency analysis. Here's what AI strategy looks like when it's not just frameworks.
No fancy GPU? No coder laptop? No problem. AI agents run fine on an old Dell XPS with a busted GPU. The heavy lifting is done by ChatGPT or Claude. Your day-to-day computer is already enough.
Why business teams need decision systems instead of passive dashboards, with a focus on decision support, operational execution, and business impact.
A business process redesign view of eliminating 120+ hours of manual reporting through automation strategy, ROI measurement, and workflow redesign.
A practitioner story on building a six-agent AI operating system: architecture, lessons learned, failures, tradeoffs, and why orchestration matters more than demos.
How AI-enabled forecasting improves planning decisions by integrating signals, operating context, and decision cadence, including a 90%+ accuracy journey.
An industry analysis of AI-native business teams, human plus AI collaboration, agent ecosystems, organizational design, and future operating models.
A business executive perspective on AI, decision quality, execution, change management, and why commercial strategy matters for applied AI.
What 90%+ forecast accuracy means in commercial planning: assumptions, executive adoption, pricing signals, pipeline data, and operational context.
A cross-industry view from Dell Technologies, Asociacion Chilena de Seguridad, and Moller y Perez-Cotapos.
How Felipe Postigo Gonzalez built AI and agentic workflow systems using Hermes / OpenClaw for coordination, reporting, analytics, and workflow execution.
A lived-experience article on Dell Technologies server revenue operations, $16B+ revenue scope, 90%+ planning accuracy, AI-enabled decision systems, and executive visibility.
How Felipe Postigo Gonzalez transformed a resource-intensive 2-person, 80-hour-per-week process into a scalable 1-person operation through business intelligence, AI-enabled workflows, and operating-system thinking.
How MBA strategy training, enterprise revenue operations, and hands-on systems building shaped Felipe Postigo Gonzalez’s applied AI profile.
A practical article grounded in the Dell MBA internship pricing framework and later server revenue operations work.
How automated forecasting and financial modeling at Moller y Perez-Cotapos improved forecast accuracy by 50% and reduced execution time by 25%.