Tasos Pardalis

@TheChelseaForum Hello

🚨 If you have less than 258k followers! 🔥 Just say "Hello" we follow you. 💯💙🫡

When you want to tell a colleague that the work they spent weeks on is useless but you have to be professional

Is PSG considered the best because they hadn’t played against @ChelseaFC before? 😜 Up the Chels 🎉🎉

Build a Storyteller & Title Guru: AI-Agent Playground with Flowise medium.com/road-to-full-s… @FlowiseAI @Medium #AIAgent #MultiagentAI #FlowiseAI #promptengineering

The Secret Sauce Behind Smart AI Agents: APIs Explained medium.com/road-to-full-s… #AI #API #MCP #ModelContextProtocol #AIAgents

I just published From Prompt Monkey to Goal Machine: AI Agents 101 medium.com/p/from-prompt-… I genuinely want to know what tool do you like to use to build your AI-Agents and why. #AIAgents #AI

If you are just starting to get curious about LLMs and RAG this might be the post for you. If you are already familiar/expert, this might be a post you could hate on 🙃 Either way, reads are valued

I just published Making Sense of RAG: Vector DBs & Embeddings Explained medium.com/road-to-full-s… #RAG #RAGmodels #LLM #vectordatabase #embeddings #vectorembeddings

@ChelseaFC I think CFC has been the better team for majority of the game but don’t understand how we lost shape after scoring. Also the three games I’ve watched so far in the tournament give me the impression that referees are biased against CFC for some reason. Or am I biased?🤔

Thoughts on the game so far, Blues? 🔴 0-1 🔵 [85]

@ChelseaFC Any updates? Are they coming back soon or should I go to bed?

The match has been suspended due to extreme weather. ⏸️ 🔴 0-1 🔵 [85]

@RJamesSZN Renato Veiga, by far the best performer of the three in the PL. Also, Cucu needs a backup :D

From my studies and experience I believe that the reason that data integrity is an afterthought is because data scientists don't learn about CI/CD/CT practices. I truly believe that learning about CI/CD/CT practices will help teams to productionise more models.

I have a masters degree in data science for business and worked as a data scientist for 4.5 years before transitioning to a data engineering role 2 years ago (my goal is to be a competed ML engineer)...

I am writing a post about RAG models and I remembered a thought I had a few months ago when I worked with a data science team to help them put a RAG model into production in a highly regulated sector. The thoughts is "Why is data integrity always an afterthought in DS/ML teams?"

@karpathy Thinking of token limits, could that be relevant to the context engineering idea?

+1 for "context engineering" over "prompt engineering". People associate prompts with short task descriptions you'd give an LLM in your day-to-day use. When in every industrial-strength LLM app, context engineering is the delicate art and science of filling the context window with just the right information for the next step. Science because doing this right involves task descriptions and explanations, few shot examples, RAG, related (possibly multimodal) data, tools, state and history, compacting... Too little or of the wrong form and the LLM doesn't have the right context for optimal performance. Too much or too irrelevant and the LLM costs might go up and performance might come down. Doing this well is highly non-trivial. And art because of the guiding intuition around LLM psychology of people spirits. On top of context engineering itself, an LLM app has to: - break up problems just right into control flows - pack the context windows just right - dispatch calls to LLMs of the right kind and capability - handle generation-verification UIUX flows - a lot more - guardrails, security, evals, parallelism, prefetching, ... So context engineering is just one small piece of an emerging thick layer of non-trivial software that coordinates individual LLM calls (and a lot more) into full LLM apps. The term "ChatGPT wrapper" is tired and really, really wrong.

I just published LLM Function Calling medium.com/road-to-full-s… Do you think function calling is a powerful tool for LLMs or not so special? #AI #LLMs #LLMfunctions #functionCalling

I just published The World of LLMs: Foundations, Models, and Practical Considerations medium.com/road-to-full-s… #AI #LLM #ML #CNN #RNN #DeepLearning

Last week I published Intro to AI-Agents medium.com/road-to-full-s… - My ambition is to write a series of post on AI-Agents as I am getting to understand them better and eventually build one or two. #AIAgent #LLM #AI

Fed up with @ChelseaFC performances the last 10 games. Even when we won we didn’t really perform. No ST is ridiculous.

@amadoit__ Honestly, doesn’t make sense. It was a clear red. And he almost scored a few moments ago.

don't see how Brooks escaped a sending off for this. recommendation was a red for violent conduct, referee has watched this and decided to give Brooks a ...yellow? a farce, as usual.