What AI Can't Do (But Python Can Do Instead)

After using AI for management deck prep and AI-generated Python code for a while, I came to a few conclusions: Using Python code from AI instead of the AI itself adds consistency to reporting. AI is non-deterministic, which makes it unreliable for this. I tried having AI generate multi-page prompts, thinking that would produce consistency, but it does not. At least not enough for me. Also, I still like PowerPoint presentations, and Claude is sometimes clumsy. Table borders spill over, fonts are unreadable, etc. Python code, however, is not flexible. To add one more slide, you need to have AI modify the code. Having the code change every reporting period is costly, both time-wise and token-wise. Best practice seems to be having a standard set of slides in the deck, then using AI for deeper dives specific to the period. I wonder how to merge the consistency of the Python approach with the flexibility of AI tools. I am told Streamlit is a good option, but I need to do more research before starting to use it. #CFO #FPandA #AIinFinance #FinanceAutomation #Python

Day 2/30 – M4Ace AI/ML Challenge One thing I learned today: Python basics are not “basic” — they are foundational to AI. If you're starting AI/ML, here are 3 core Python concepts you must understand: 🔹 Variables & Data Types Everything in AI starts with data—numbers, text, or categories. Python helps you store and manipulate them efficiently. 🔹 Lists (Data Handling) Lists allow you to group data together. In machine learning, datasets are often handled as structured collections like this. 🔹 Functions (Reusability & Logic) Functions let you write clean, reusable code. This becomes critical when building models and data pipelines. 👉 Why this matters: Machine learning is not just about algorithms—it’s about how you prepare, structure, and process data before the model even begins. For me, this is already connecting to telecom: Network data (traffic, latency, users) must first be structured properly before any intelligent decision can be made. Strong foundation → Better models → Smarter systems. #M4ACELearningChallenge #LearningInPublic #AI #Python #Telecom

Free way to use AI models. OpenRouter a great platform for developers who want access to multiple AI models using a single API key. In this blog, I covered: * How to sign up and log in * How to create API keys * How to explore free AI models * Python code examples to get started * Why OpenRouter is useful for developers building AI projects If you're learning AI integration, building side projects, or exploring LLM APIs, this guide will help you get started quickly. Would love to hear your thoughts and experiences with OpenRouter. #minfy #OpenRouter #AI #Python #LLM #APIs #MachineLearning #Developers #TechBlog #Programming #GenerativeAI

Day 2/30 – M4Ace AI/ML Challenge One thing I learned today: Python basics are not “basic” — they are foundational to AI. If you're starting AI/ML, here are 3 core Python concepts you must understand: 🔹 Variables & Data Types Everything in AI starts with data—numbers, text, or categories. Python helps you store and manipulate them efficiently. 🔹 Lists (Data Handling) Lists allow you to group data together. In machine learning, datasets are often handled as structured collections like this. 🔹 Functions (Reusability & Logic) Functions let you write clean, reusable code. This becomes critical when building models and data pipelines. 👉 Why this matters: Machine learning is not just about algorithms—it’s about how you prepare, structure, and process data before the model even begins. For me, this is already connecting to telecom: Network data (traffic, latency, users) must first be structured properly before any intelligent decision can be made. Strong foundation → Better models → Smarter systems. #M4ACELearningChallenge #LearningInPublic #AI #Python #MachineLearning #DataScience #Telecom

Over the past few weeks, I worked on building real-world AI systems and understanding how things work under the hood. Here’s what I explored: • LLM fundamentals (transformers, attention, embeddings) • Prompt engineering techniques • Building RAG pipelines with LangChain • Designing AI agents using LangGraph • Implementing memory systems (vector + graph) • Running models locally with Ollama & Docker • Creating scalable AI backends using FastAPI + Redis I also built hands-on projects like: • AI RAG system for document search • CLI-based coding assistant • Voice-enabled AI agent Also building Promptix : an AI SaaS platform for: • Chatting with PDFs (RAG) • Resume analysis (ATS + insights) • Mock interviews & prompt tools This journey helped me move beyond just “using AI” to actually building AI systems. Next step: Applying these skills to real-world products. #AI #MachineLearning #LLM #LangChain #GenerativeAI #Python #SoftwareEngineering

To start an AI learning journey, there’s one place to begin: Python 🐍 One of the most practical, no-fluff resources available is by . No hype. Just clarity. Here’s why it stands out 👇 ▶️ Starts from zero Variables, data types, operators, syntax — all explained cleanly without overwhelm. ▶️ Logic-first approach Conditionals, loops, and functions taught in a way that actually makes sense. ▶️ Core data structures done right Lists, Tuples, Dictionaries, slicing — the building blocks of real-world data work. ▶️ Ends with real capability Concepts are not just introduced — practical coding becomes possible. 💡 Python remains the #1 language for AI and data science. The starting point doesn’t need to be complicated. This is it. Follow for practical AI and engineering resources. Repost so more builders can get started 🚀 Follow and Connect: Woongsik Dr. Su, MBA #Python #AI #DataScience #MachineLearning #Programming #LearnToCode #CodingForBeginners #Analytics #TechSkills #AIJourney

Everyone wants to get into AI. Almost no one knows where to start. So let’s clear the noise: 👉 If you’re starting today, pick Python. No debate. Not because it’s “easy”… But because it lets you build faster than you can overthink. While others are stuck choosing between 10 languages, you could already be: Training your first model Playing with real datasets Building something people can actually use That’s the real game. Python isn’t just a language — it’s the entry pass to the AI ecosystem. From machine learning to deep learning to automation… everything connects back here. But here’s the truth most people ignore: ❌ AI is not about coding ✅ AI is about thinking in data, patterns, and problems The language just helps you express it. So don’t aim to “learn Python” Aim to build with Python Start messy. Stay consistent. Ship small wins. Because in AI, the winner is not the smartest person— It’s the one who started early and didn’t stop. What’s stopping you from starting today? #machinelearning #ai

🚀 Day 7 of My Generative & Agentic AI Journey! Today’s learning was quite interesting — I explored Operator Overloading and Bytearray in Python. Here’s what I learned: ➕ Operator Overloading: Operators behave differently depending on the data type. • Example with numbers: 2 + 3 = 5 • Example with lists: A = [23, 56, 45] B = [1, 2] A + B = [23, 56, 45, 1, 2] • Another example: C = ["apple"] C * 3 = ["apple", "apple", "apple"] 👉 Same operator, different behavior — that’s the power of Python! 🔤 Bytearray: • Learned about bytearray — a mutable sequence of bytes • Unlike strings, bytearray can be modified • Example: b = bytearray(b"hello") b[0] = 72 → converts ‘h’ to ‘H’ Result: bytearray(b"Hello") 🔄 Replace in Bytearray: • We can modify content using methods like replace() • Example: b = bytearray(b"hello world") b = b.replace(b"world", b"python") Result: bytearray(b"hello python") 👉 Key takeaway: Understanding how Python handles operations and low-level data types like bytearray gives more control over data processing. Learning something new every day and going deeper 🚀 #Day7 #Python #GenerativeAI #AgenticAI #LearningJourney #BuildInPublic

Learning Python today is no longer just about syntax. It’s about enabling systems that can think, decide, and act. With the rise of Agentic AI, the role of Python is evolving rapidly. It’s not just a programming language anymore. It’s becoming the foundation for building intelligent, autonomous workflows. ⸻ 🧠 What Makes Agentic AI Different? Unlike traditional systems: • It doesn’t just execute instructions • It can plan tasks • It can choose tools • It can adapt based on context • It can take multi-step actions ⸻ ⚙️ Where Python Fits In Python enables this ecosystem by making it easier to: ✔ Integrate with LLMs and AI models ✔ Build orchestration layers for agents ✔ Connect APIs, tools, and data sources ✔ Prototype and scale intelligent workflows ⸻ 🔍 The Real Learning Shift It’s no longer just: 👉 “How do I write this function?” It’s becoming: 👉 “How do I design a system where an agent can solve this problem?” ⸻ 🚀 As an Integration Architect, This Feels Like a Big Shift We are moving from: • Static workflows → to • Dynamic, AI-driven systems Where integration is not just about connecting systems… But enabling intelligent interactions between them. ⸻ 🔥 Final Thought Agentic AI + Python is not just a new skill. It’s a new way of building software. ⸻ What’s your experience so far with Agentic AI — learning, experimenting, or using in production? ⸻ #AgenticAI #Python #AI #SoftwareArchitecture #IntegrationArchitecture #LLM #FutureOfTech #TechLearning

🚀 Why Python is the Backbone of AI & Machine Learning Over the past few months, while working on projects like a Student Dropout Prediction model and a RAG-based Document Q&A system, one thing became very clear — Python is not just a programming language, it’s an ecosystem powering AI innovation. Here’s why Python stands out for AI/ML 👇 🔹 Simplicity & Readability Python’s clean syntax makes it easier to focus on solving problems rather than writing complex code. 🔹 Powerful Libraries From data processing to advanced AI models: • NumPy & Pandas for data handling • Scikit-learn for machine learning • TensorFlow & PyTorch for deep learning • OpenAI & LangChain for Generative AI 🔹 Strong Community Support Python has one of the largest developer communities, which makes learning, debugging, and building faster. 🔹 End-to-End Capability From data collection → preprocessing → model building → deployment — Python supports the entire AI pipeline. 💡 In my recent projects: • Built a Machine Learning model to predict student dropout risks with high accuracy • Developed a RAG-based system to answer questions from documents using LLMs These experiences reinforced how powerful Python is in turning ideas into real-world AI solutions. 📌 If you’re starting your journey in AI/ML, Python is the best place to begin. #Python #AI #MachineLearning #DataScience #GenerativeAI #LLM #OpenAI #LangChain #CareerGrowth #knowledgetransfer