Practice

DeepSeek has once again followed its “quiet release” strategy, making its new DeepSeek-V3-0324 model available on Hugging Face without any major announcements. Instead of marketing hype, they’ve simply delivered a solution for the community to evaluate. I tested the model locally on a Mac Studio equipped with an M3 Ultra chip and saw impressive performance, generating over 20 tokens per second. This marks a significant acceleration for running capable models on local hardware, making it a viable option for developers. ...

I wanted to share what the final stretch before launching an AI product feels like. With just 10 days until AVELIN goes public, the pressure is immense, and the reality is a mix of controlled chaos and sharp focus. Our team is feeling the strain. As tasks pile up and deadlines loom, fatigue is a real factor. We’re constantly fighting the dilemma of adding ‘just one more thing’ versus locking in the release date and moving forward. ...

Today, we are officially launching AVELIN—the Artificial Intelligence my team and I have been building for the last three years. Our journey began with humble pilots, experimenting with the first GPT models and running foundational tests. We quickly evolved from simple, single-model chatbots to developing our own proprietary training system, complete with knowledge ingestion, document storage, and our first implementations of Retrieval-Augmented Generation (RAG). ...