Blog.
My musings on anything from leadership to technology to entrepreneurship and back.
AI for Wordpress Development
Here's a real world example on how gpt-4o can help you writing plugins in Wordpress. Massive time saver.
B2B Startup Success Secrets
Martin Tobias, a seasoned pre-seed venture capitalist, offers crucial insights from 250 investments in B2B startups. Incredible wisdom. Let's explore!
Interim CTO vs. Fractional CTO
Fractional CTO and Interim CTO - same same or different? And what do these differences mean for you as company that needs an expert to fill the leadership gap?
The Best AI Tools
An updated list of the best AI tools. The landscape is changing fast. Bookmark the page to stay up to date!
Fine-Tuning a Chat GPT AI Model LLM
Let's explore how simple it is to fine tune an existing model using OpenAI's Chat GPT models. But maybe - just maybe - a RAG is the better approach to get started with custom LLMs. Let's explore!
On Modular Monoliths
When it comes to microservices I am always reminded of Martin Fowler's quote: '...you shouldn't start with a microservices architecture. Instead begin with a monolith, keep it modular, and split it into microservices once the monolith becomes a problem.' Let's check out how we can put this into action.
Privatracker Tech Stack
Privatracker is a SaaS tool designed to serve as a privacy-respecting alternative to traditional web analytics platforms like Google Analytics. Let's check out the tech stack and discuss approaches and alternatives.
RAG LLM - Meta Llama 3.2 vs OpenAI GPT-4o
Retrieval-Augmented Generation (RAG) offers a great way to provide data to Large Language Models (LLMs) and generate responses based on your own content. Using private LLMs like Meta's Llama3 3.2 is particularly beneficial when dealing with sensitive data. But what if your use case doesn't involve sensitive data? Can a publicly accessible model like OpenAI's GPT-4o improve the outcome? Let's compare!
Improving RAG LLMs
In my previous blog post I experimented with a locally running the Llama32 model to answer questions using my potentially sensitive content. While it performed okish, I wasn't completely satisfied with the results. Let's see if we improve that.
Custom LLM with a RAG
Recently, I've been working on tech diligence and strategy projects, which involve quickly accessing scattered information from various reports and documents. Understanding company products, roadmaps, and customer segments is crucial, and consolidating this data into a single source is a good start, but often not enough. Large language models (LLMs) offer a solution by allowing you to ask direct questions. I implemented a pipeline that securely processes sensitive data and answers questions comprehensively. Let's check it out.
