Engineers Are 10x Faster with AI, Why Aren’t We Seeing 10x More SaaS?
No AI used in this post. Other posts may contain up to 10% AI assistance.
Mehmet Burak Sayıcı
This is about me
Latest posts
How to design a RAG System : A Boilerplate
🎥 Watch the demo on Vimeo GitHub Link A RAG system that is : Chunking with propositional model + late chunking, and simple recursive overlap retrieval Using Qdrant as a Vector DB, utilising its hybrid search (BM25 + Dense Search) Reranks via LLMs Using query enhancement agent via LLMs Using crewAI for conversation+retrieval agent Allows you to chat with data Dividing conversations into sessions,…
Clever Chunking Methods Aren’t (Always) Worth the Effort
I was reading about fancy chunking strategies on RAG systems. There are endless ways of chunking which most of them makes sense. Reading through it, I’ve even seen technical blog posts/papers that mentions 25 ways of chunking?!? Then I realised it’s not about the fanciness of the methods but how much performance you gain by an effort. Lots of readings/podcasts of Frank Liu (former CTO of Zillus,…
Information Retrieval Fundamentals #1 — Sparse vs Dense Retrieval & Evaluation Metrics: TF-IDF, BM25, Dense Retrieval and ColBERT
Information Retrieval could be defined as a field of study interested in ways to finding relevant material of an unstructured nature (usually text) from a collection of data (i.e.: files stored in your file system) to satisfy user information need.
LLM-SQL Tools
I reviewed SQL-LLM-Agent tools and libraries, sorted by GitHub stars. I just searched repo tags matching #llm #sql and similars.
LLM Interviews : Hosting vs. API: The Estimate Cost of Running LLMs?
I’ll use any resource that helps me to prepare faster. I’m not directly working for interviews, but preparing for them makes me learn so many stuff, for years(check main page of website for 8 yo youtube channel). I’ll cite every content I use. I’ll agressively reference anything I believe explained things better than me.
LLM Interviews : Prompt Engineering
I’ll use any resource that helps me to prepare faster. I’m not directly working for interviews, but preparing for them makes me learn so many stuff, for years(check main page of website for 8 yo youtube channel). I’ll cite every content I use. I’ll agressively reference anything I believe explained things better than me.
LLM Interviews : Vector DBs
I’ll use any resource that helps me to prepare faster. I’m not directly working for interviews, but preparing for them makes me learn so many stuff, for years(check main page of website for 8 yo youtube channel). I’ll cite every content I use. I’ll agressively reference anything I believe explained things better than me.
GPT1 Implemented with NumPy : Generate Jokes
I love implementing DL algorithms with NumPy, it’s best way to learn the engine. I’ve implemented SHAP, and CNNs, and GradCAM methods with pure numpy before, you can refer to my Github Page for this.
Easy-to-use NoSQL Prompt Database for Small Projects
Ever needed a database for your LLM project that: