The AI landscape has been taken by storm in the last dozens of months with the spread of important new technologies and paradigms. During the talk we will explore the steps involved in building an NLP project starting from the raw data until the end product, summarizing what we have learned in creating a complete end to end pipeline involving data anonymization, data exploration, embedding and vector databases and fine tuning. The last goal of this talk is to run through and share the learnings and the fun of building with these magical open source tools.