AAU Software Engineering community

✨Quick tip: Don't just paste your bug into ChatGPT/Claude paste the right context ✨ Most people get bad AI debugging answers because they only paste the error message. Instead, give it: - The error/traceback (obviously) - The actual function, not just the line that failed - What you expected vs what actually happened - Anything you already tried (saves it from suggesting the same dead end) One extra move that changes everything: ask it to explain why the bug happens before asking for the fix. If the explanation doesn't make sense, the fix probably won't either and you catch a wrong AI answer before you ship it. Hope this helps ☺️

Quick tip: Don't just paste your bug into ChatGPT/Claude paste the right context Most people get bad AI debugging answers because they only paste the error message. Instead, give it: The error/traceback (obviously) The actual function, not just the line that failed What you expected vs what actually happened Anything you already tried (saves it from suggesting the same dead end) One extra move that changes everything: ask it to explain why the bug happens before asking for the fix. If the explanation doesn't make sense, the fix probably won't either — and you catch a wrong AI answer before you ship it. 💬 What's the worst "confidently wrong" fix an AI ever gave you?

5 Things Every Software Engineering Student Should Understand Before Graduating You don’t need to know everything, but you should understand these: How the Internet actually works DNS, HTTP/HTTPS, APIs, servers, databases, and what happens when you visit a website. Git & GitHub Branching, commits, pull requests, merging, conflicts, and collaborating with other developers. Databases SQL, relationships, indexes, transactions, normalization—and when NoSQL makes sense. Software Architecture How to structure a real application, separate responsibilities, design APIs, handle errors, and think about scalability. How to Read & Debug Code You won’t always write code from scratch. A huge part of engineering is understanding existing code, finding bugs, and fixing them. Bonus: Learn how to use AI coding tools without becoming dependent on them. Understand the code you ship. Your degree teaches you concepts. Your projects teach you how to build. #softwarecheck @CTBEsoftwareengineers

Something valuable is coming your way…

Hello family of CTBESoftwareEngineeerrs 👋 We’re starting something new weekly, worth-reading articles made just for you.

Each week, expect content focused on:

• Software & technical skills • Career growth • Lessons & real experiences 📅 Starting September 1 (2019 EC). Make sure you don’t miss the first drop it might be the one that changes how you think or build. Stay tuned 🔥 @CTBEsoftwareengineers

Applications are now OPEN for the Claude Campus Ambassadors Program! This year, the program is expanding with three tracks:

• Undergraduate students • Graduate students • PhD/Postdocs

If you're a student at Addis Ababa University (AAU) and passionate about AI, tech, and building communities, this is a great opportunity to get involved, grow your network, and represent your campus. ⏳ Deadline: [September 12 2026 ] 🔗 Learn more & apply: [ CLICK HERE ]

https://meet.google.com/qdi-xwwi-uku Use this guys

Do you remember the first time you deployed your frontend and backend using Render and Vercel? It felt like one small mistake, like missing an environment variable or using the wrong build setting, could break everything and you would not even know where to start. But after doing it again and again, facing failed builds, random 404 errors, and fixing configurations step by step, you reach a point where you are no longer scared. Now you are confident enough to debug whatever goes wrong and get your app running.

Here is what your life look like when you're software Engineer

1. Write code once → you understand it. 2. Write it 20 times → you remember it. 3. Build 10 projects → you start seeing patterns. 4. Debug 100 errors → you stop panicking when something breaks.

Repetition changes:

• Syntax → muscle memory • Bugs → debugging skills • Projects → problem-solving • Docs → faster learning • DSA → pattern recognition • APIs → better system understanding • Deployments → confidence • Failures → better engineering decisions Build → Break → Debug → Fix → Repeat. @CTBEsoftwareengineers

🇰🇷 South Korea Makes History: First G20 Nation to Offer Free AI to All Citizens South Korea's Ministry of Science and ICT has selected three consortia led by SK Telecom, KT, and Kakao to build a free, unlimited AI chatbot and public-service assistant for all 52 million residents. 📌 Key details 🔹 Backed by 512 Nvidia B200 GPUs from the government 🔹 At least 50% of usage must run on Korean-made foundation models (EXAONE, Solar, Kanana, A.X K2, Motif, etc.) 🔹 Basic chatbot + public-service agent will be free with no usage limits 🔹 Beta launch expected soon, full rollout by year-end 🔹 Part of a $7.2B national AI investment push in 2026 This makes Korea the first G20 country to treat AI access as a public service aiming for an "AI-basic society" regardless of income, age, or digital literacy. #TechNews @CTBEsoftwareengineers

Moving forward, we will be conducting weekly performance reviews of our admin team. Those who are not actively contributing will have their admin rights revoked and we will be looking to promote new, dedicated admins who can consistently meet the interest of this community. Those who have creative ideas to contribute to this community make yourselfs ready not because we pay you but because of your passion to share some knowledge or adding value for others

AI token prices are hitting record low and that's mainly because of the cheap Chinese models

📢 GLOBAL TECH WEEKLY: Late August 2026 Dev Roundup If you are building your portfolio or preparing for tech interviews, here are the massive global updates from this week that you need to know. 👇 1. The AI Tool Landscape Shifts GitHub Copilot Gets Flexible: On August 28, Copilot rolled out a major update that finally lets developers choose which underlying AI model they want to use for reasoning. It also introduced new features that allow teams to share specialized, custom AI agents across their workspaces. The AI "Hidden Tech Debt" Crisis: A massive new industry report highlighted a side effect of everyone using tools like Cursor and Copilot: copy-pasted code has exceeded refactored code for the first time in history. Tip for students: Companies are drowning in unreviewed AI code. Focus on learning how to architect, test, and safely review AI-generated code to stand out. 2. The Next Wave of Models & Agents OpenAI’s GPT-5.6 Era: OpenAI's new GPT-5.6 family (split into Sol, Terra, and Luna tiers) is now fully deployed, alongside ChatGPT Work—an agentic system capable of executing complex, multi-hour engineering projects across team files. Agentic Engineering is the New Standard: Industry analysts noted this week that "Agentic AI" is officially crossing from fun experiments into production environments. Future roles won't just ask you to "prompt" an AI, but to build scalable platforms that orchestrate these agents. 3. Infrastructure Reality Checks The 8-Hour GitHub Outage: If your commits or pipelines failed on August 17, you weren't alone. GitHub and Copilot suffered a massive global outage that knocked out Actions, APIs, and Pull Requests for nearly a full workday. A great reminder for DevOps and Cloud students: always have fallback plans and avoid single points of failure in your CI/CD pipelines. @CTBEsoftwareengineers #News

Project Showcase | Group Pilot Group Pilot is a Telegram-based team collaboration bot designed to simplify how teams manage work and stay organized inside group chats. ✨ Key Features: • Create and assign tasks to team members • Track task progress in real time • Smart reminders for deadlines and updates • Easy team collaboration inside Telegram • Automatic notifications to keep everyone aligned Group Pilot helps teams reduce confusion, stay productive, and move projects forward smoothly — all within Telegram. Built with: Python, FastAPI, Docker & Google Cloud Platform (GCP) Bot: @om_groupPilot_bot @CTBEsoftwareengineers #project_showcase

Calling all Creatives, Hackers, & Innovators in Addis! Are you ready to collaborate, build, and bring your creative ideas to life? Join us at Addis Jam for a unique hackathon experience designed for creators, developers, and artists! 📍 Location: Gudina Tumsa Conference Center 🔗 Register Now: https://forms.gle/TVvGHvvbm4wu6vuW9 Don't miss out on connecting with the local creative tech community. Secure your spot today! 🔥 addisjam #creativecommunity #ethiopiancreatives #creativehackathon #arthackathon

One mistake many people make when applying for remote jobs is applying without direction.

 They apply for every remote job they come across without knowing exactly what role they want. As a result, their applications become unfocused, and their CV doesn't clearly communicate what they're best suited for. When you don't know what you want, it's easy to spend time applying for jobs that don't match your skills or interests. After facing several rejections, some people begin to believe that getting a remote job is impossible.

Before you start applying, be clear about the type of remote job you want.

Don't just say, "I want a remote job." Ask yourself: What role am I looking for? What skills do I have? Which jobs match those skills? The clearer you are about the role you want, the easier it becomes to tailor your CV, write stronger job applications, and apply for jobs that match your skills.

Being clear about the role you want doesn't guarantee that every application will get accepted, but it will help you apply with more purpose instead of applying for every job you see.

@CTBEsoftwareengineers

How does Database Replication actually work? Database replication is basically keeping copies of the same database on multiple servers. Instead of relying on one database for everything, changes can be replicated to other databases, allowing the system to handle more reads and continue working if one server fails. But replication introduces its own problems like replication lag, where a replica hasn't received the latest changes yet, and the question of whether data should be replicated synchronously or asynchronously. There are also different replication approaches, such as primary-replica and multi-primary replication, each with different trade-offs. https://www.frontendtechlead.com/learn-system-design/database-replication

Sharing this because I found it genuinely useful. I was using some of those “free” platforms like Render for my projects, but the limitations started becoming frustrating especially when a free database could be deleted after a month. I actually ended up losing data from some of my apps because of that. That’s when I started looking more into GCP. It gives you access to actual cloud infrastructure, so instead of just deploying an app, you can manage things like your backend, database, storage, and VMs in one place. And the good part is that GCP gives new users $300 in credits for 3 months You can use those credits for things like Cloud Run to deploy backends, Compute Engine for VMs, Cloud SQL/Firestore for databases, Cloud Storage for files, and more. Google

What happens when one microservice starts failing? Imagine your API depends on a Payment Service. Everything works normally, until the Payment Service becomes slow or starts returning errors. Your API keeps sending requests anyway. Each request waits for the Payment Service to respond, so connections and resources start getting tied up. As more requests arrive, the problem spreads. The Payment Service is struggling, but now your API is struggling too. This is where the Circuit Breaker Pattern comes in. Instead of repeatedly calling a service that is clearly failing, your system can temporarily stop those calls and fail fast. But how does it know when to open the circuit, when should it try the service again, and what happens when the service recovers? Circuit Breaker Pattern System Design Academy @CTBEsoftwareengineers