Major League Hacking

Congrats to this week's top 7 authors! 🎉 🔹 Deepu K Sasidharan — built a fully offline local AI-assisted Linux dev machine 🔹 Harlimat Odunola — made AI agent training verifiable by putting every episode on-chain with Solana 🔹 Debbie O'Brien — documented an entire product in four days using an open-source AI agent 🔹 Ken W. Alger — engineered agent memory as a true architectural discipline with working, semantic, and episodic layers Read these and more: https://lnkd.in/ed-3X2KR

Want to learn Solana and share what you learn along the way? We’re running a 100 Days of Solana Writing Challenge on DEV, from 15 May to 22 May. 100 Days of Solana is a free learning challenge for web and mobile developers who want to get hands-on with Solana without the hype or mystery. For the writing challenge, we’re inviting developers to publish a DEV post that helps another developer understand, build, debug, or try something in Solana. You don’t need to be a Solana expert. Useful, honest, practical learning posts are exactly what we’re looking for. You could win one of three $500 prizes: • Most helpful post   • Most read, on-topic post   • Best posting streak: the strongest set of four high-quality, helpful posts during the challenge week  There are also ten DEV++ subscriptions for notable submissions. To take part: 1. Register for 100 Days of Solana: https://lnkd.in/eqcAmW26 2. Publish your post on DEV between 15 May and 22 May 3.Include the 100daysofsolana tag If you’ve been curious about Solana but haven’t known where to start, this is a good moment to jump in, learn something concrete, and share it with other developers. Full writing challenge details: https://lnkd.in/eHBbpjHj #100DaysOfSolana #Solana #Web3 #DeveloperEducation #LearnInPublic

What does it take for a developer trust your docs the first time they read them?
 Danielle Washington has spent 11+ years answering that question across some unlikely terrain: teaching English in Spain, founding a brand, shipping Kubernetes tooling for one of the biggest open-source vector databases in the world. The through line? The instinct to find the gap between what a system does and what its users can actually understand, then close it.
 At HPE, that meant architecting developer portals for GreenLake. At Weaviate, owning deployment documentation end-to-end and building tooling to reduce EKS friction. Now the same approach is being applied to AI and information retrieval at 10aLabs.
 Work that lives where documentation, developer experience, and product thinking converge. See Danielle W speak at DevRelCon NYC this July 22-23.
 Register at https://nyc.devrelcon.dev/

Join the Hermes Agent Challenge! 🚀 If you've been watching the open source AI space, Hermes Agent has probably been on your radar. If not, now's the time to get acquainted. Built by Nous Research to run on your own infrastructure, it's capable of planning, tool use, and multi-step reasoning across complex tasks. The kind of thing worth digging into firsthand. Running through May 31, this challenge has opportunities for both builders and writers: 🟣 Build With Hermes Agent: automate a workflow, spin up a research pipeline, create a multi-tool assistant, or experiment with something entirely your own. The scope is wide open, as long as Hermes Agent is doing real work at the heart of it. 🟣 Write About Hermes Agent: tutorials, comparisons, opinion pieces, deep technical breakdowns. If it educates, inspires, or sparks discussion, it qualifies. 🟣 $1,000 in prizes. 8 winners total (4 per prompt). Every valid submission also earns a DEV completion badge. Submissions close May 31 at 11:59 PM PDT. Winners announced June 4. Good luck and happy coding! https://lnkd.in/e2kpmjpX

Can you gamify... showering? Meet Ranked Showers, an ingenious (and hilarious) hackathon project by Alexander Gu, Ryan Chang, Inuka Silva, and Andrew Whitehead. The team wanted to tackle the "smelly dev" stereotype by creating a leaderboard system that encourages computer scientists and engineers to stay fresh. This isn't just a simple timer—it's a full-stack hardware and software solution. The team used temperature, humidity, and distance sensors to monitor the environment, then took it a step further by using a Gemini model to analyze audio from AirPods to verify the user is actually showering. To top it off, they brought the data onto the blockchain. Using MongoDB for storage and Solana for rewards, they minted custom coins to reward the "cleanest" builders and those practicing eco-friendly habits. It’s a perfect example of how builder culture can take a fun idea and back it up with serious technical execution! Read more about their build here: https://lnkd.in/emYZYXV6

Urban pollution is a massive challenge, but the solution might already be hanging over our heads. 🚦🧹 Check out StreetSweep AI, a project built by Lukhsaan Elankumaran, Andrew Law, and Varun Gande that uses artificial intelligence to map and manage city litter in real-time. By integrating public traffic camera feeds with the Gemini API, StreetSweep AI automatically identifies trash buildup and ranks it by severity. Combined with user-submitted images, the platform creates a "Volunteer Dashboard" that helps city planners and cleanup crews visualize exactly where help is needed most. It’s an incredible example of how builders can take modern technology—like AI and MongoDB—and apply it to decades-old infrastructure to solve real-world problems. Projects like this are exactly why we celebrate the hacker community! Read more about their build here: https://lnkd.in/eaQNPXMa

Bring your 100 Days of Solana questions!

Most growth teams spend 80% of their budget trying to capture demand. Matt Henderson spends 70% of his creating it. As Director of Growth at Sentry, Matt has built one of the more unconventional playbooks in B2B: developer podcast acquisitions, Warriors and F1 sponsorships, heavy investment in influencers and OOH, all in the name of brand awareness over attribution-friendly channels. Matt’s take: the channels that build authentic brand presence with developers are now the same ones LLMs draw from when making product recommendations. Brand awareness and AI discoverability are no longer separate problems. We're excited to welcome Matt as one of our speakers at DevRelCon this July 22-23. Register at nyc.devrelcon.dev

Congrats to our top 7 authors! 🎉 This week:  🖥️ Jeongho Nam's wild VR multi-agent coding setup  🤖 Rohini Gaonkar breaks down AI hallucinations  💡 Ros Sopheak • Adam - The Developer on why we should write code that is easy to delete 💥 Fernando Fornieles builds an app once by hand and again with AI, then compares results Plus more! 

Arc 4 of #100DaysOfSolana starts today, and we’re shifting from using Solana to understanding how it actually works underneath. So far, we’ve covered: • wallets • reading blockchain data • transactions • confirmations Now we’re going deeper into the system behind all of it: Solana’s account model. This arc explores: 🔍 inspecting real on-chain accounts 🛠️ building a Solana account explorer 🧩 decoding raw account data ⚙️ exploring system programs and sysvars 🌐 navigating Solana Explorer like blockchain DevTools ✍️ explaining the account model in plain English for Web2 developers If you've ever wondered how Solana actually stores state, permissions, balances, programs, and data under the hood, this is the part of the journey where it starts to become concrete. One of the biggest mindset shifts for Web2 developers is realizing that on Solana, everything is an account: • wallets • programs • tokens • NFTs • application state This arc is designed to make that model feel intuitive instead of abstract. Join free: https://lnkd.in/eqcAmW26