ፊሊጶስ (@philipbuilds)

AI (Artificial Intelligence) is technology that enables computers to learn, reason, recognize patterns, and perform tasks that normally require human intelligence. 🤖🧠 From ChatGPT and voice assistants to recommendations and self-driving technology, AI is already part of everyday life…

One career lesson that becomes clearer with experience: your career is a long game. Working hard matters, but hard work alone doesn't guarantee growth. The real advantage comes from continuously building skills that compound over time. Learn technical skills. Develop communication skills…

Building two systems simultaneously with Claude Code 🦾💻 AI-assisted development is changing how fast we can build, test, and ship. 🚀 #ClaudeCode #AI #SoftwareEngineering #Coding #AICoding

Tired of digging through massive repos just to see how LLMs work? 🤯 Stop wasting time! OpenArch is a goldmine for AI devs. It strips away all the bloat and gives you clean, hand-written, single-file PyTorch implementations of major models like Llama, Qwen, and DeepSeek side-by-side. 🧑‍💻🔥 Perfect…

Mini apps are becoming an important part of modern digital platforms. Instead of asking users to install a separate app for every service, a super app can host smaller applications for payments, loans, insurance, shopping, transport, and other services inside one ecosystem. Most mini apps run…

Why are MCP servers important? MCP servers are important because they allow AI assistants to move beyond simply answering questions. They give AI a standardized way to connect to external tools and data—such as databases, APIs, files, GitHub, Slack, or internal company systems. Think of MCP as a…

A regular day at the office, but there’s a bigger journey happening behind the scenes. 🧠💻 I’ve spent years building software, and now I’m pushing myself deeper into Applied AI Engineering. Not just learning how to use AI tools, but understanding what happens underneath — LLMs, tokens, embeddings…

🧠 Applied AI Engineering, Part 4: Understanding Embeddings How can an AI system understand that “I love programming” and “Software development is my passion” have similar meanings even though the words are different? One important piece of the answer is embeddings. 🔹 What are embeddings?…

🧠 Applied AI Engineering - Part 3: Tokens & Tokenization Have you ever wondered how an LLM actually processes the text you send to it? The answer starts with tokens. 🔹 What is a token? A token is a small unit of text that an AI model processes. Depending on the tokenizer and the text, a token can…

🧠 Applied AI Engineering, Part 2: Understanding the AI Landscape Before we start building AI-powered applications, it’s important to understand the different concepts that make up the AI ecosystem. Break down the relationship between: 🔹 Artificial Intelligence (AI) - the broader field of creating…

🤖 What is AI Engineering? AI Engineering is the discipline of building real-world applications powered by AI models. It goes beyond simply calling an LLM API. An AI Engineer needs to understand how to: 🔹 Work with LLMs 🔹 Design prompts and AI workflows 🔹 Generate and use embeddings 🔹 Build RAG…

What exactly is OpenRouter? 🤔 OpenRouter is a unified API platform that gives developers access to AI models from different providers through a single interface. Instead of integrating separately with OpenAI, Anthropic, Google, Meta, and other model providers, you can use OpenRouter to access and…

🚨 A mysterious AI model just dropped on August 20… and nobody knows who built it. Meet Ox Alpha — a stealth reasoning model with a massive 1M-token context window, multimodal input, and a serious focus on coding and AI agents. The interesting part? Its creator is still anonymous, while developers…

Finetuning LLaMA 3.1 8B to be a financial advisor for small businesses KE Next up: training on Runpod , then it's live #ai #llm #ke

Have you ever wanted to try the latest AI models but realized you'd need accounts from five different companies? GPT from OpenAI. Claude from Anthropic. Gemini from Google. DeepSeek. Llama. And every provider has different APIs, pricing, and setup. Well, today I discovered something called…

Today I discovered Microsoft's Agent Package Manager (APM) and it's an interesting idea for AI engineering teams. Think of it as npm, but for AI agents. Instead of manually sharing prompts, coding standards, MCP configurations, and agent instructions across tools like Copilot, Cursor, Claude Code…

A lot of teams underestimate how powerful Microsoft Copilot Studio has become. With little to no code, you can build AI agents that: ✅Search SharePoint, PDFs & websites ✅Answer grounded questions ✅Call APIs & automate workflows ✅Integrate with Teams, CRM & ERP systems ✅Orchestrate multiple…

we did a session today with the Microsoft team on building AI agents using their copilot studio, it's really amazing with what you can do with it #AI #copilotstudio #aiagents

full stack + AI solo dev - the orchestration layer is finally working, service bindings, dead letter cues, cron pipelines, graph traversals and model fallback chains. shipping anyway 🌃🦉 #fullstackdeveloper #AI #llms

Would you use this? AI — watch the demo & tell me what you think 👇 🌐 arifa.today 💬 t.me/ArifaTodayBot 📱 Beta on Google Play #arifa #technews #kenya #eastafrica #buildinpublic

Right now, you can build a full app in a day with AI. But that’s the problem… People are shipping: messy architectures unnecessary features code they don’t even understand AI made building easy… but understanding is now the real skill 💡 The devs winning right now are the ones who: keep things…

🧠 1. LangChain What it is: A modular framework for building LLM-powered apps and agents using chains, tools, and memory. Core idea: Break AI workflows into steps (chains) and give models tools (APIs, DBs, functions). Strengths: Massive ecosystem (plugins, integrations) Supports tools, memory, RAG…

I kinda miss Stack Overflow… Spending hours digging through threads, finding that one answer from 2013 that somehow fixed everything 😅 But now? I just ask AI anything. No tabs. No scrolling. No “marked as duplicate.” It’s faster, smarter, and honestly… a little too easy. Feels like we traded the…

POV: You stop coding alone and start using AI like a team 🤯 I use GitHub Copilot in VS Code to write code fast, then Claude CLI to review, debug, and optimize everything. Sometimes I run multiple Claude sessions at once — like having a developer, reviewer, and performance engineer all working with…

I will never regret getting this skill

Built a production-grade API 💻⚡ Cloudflare Workers makes it fast, scalable, and serverless! #WebDev #serverless #cloudflare #API #developerlife

💻 In 2017, WikiLeaks dropped Vault 7, revealing CIA hacking tools… but buried inside was something super ordinary: a Git tip! 😲 Over time, your Git repo turns into a graveyard of old branches. 🪦 Deleting them one by one? Painful. The CIA devs had a clean solution: _______________________ git…

While you’re just shipping tickets… I’m protecting the codebase from future disasters. You see extra work. I see avoided production bugs. You see “it works.” I see “will this scale?” Average devs code for today. Great devs code for the next 3 years. Raising the bar makes people uncomfortable. But…