I'm trying to land an in-house content role before the end of Q4, so I decided to take my job hunt public. Here's how I'm doing it: I research a company, identify a communication opportunity, create something useful, and pitch it to the team. For Day 2, I've done that for @atumhq First, a little about Atum. Atum is building an open payments network for global money movement. One thing that stood out to me is how clearly Atum defines what it is and what it isn't. Atum is not a blockchain, a stablecoin or a settlement layer. It is a coordination layer for money movement with a 5-step payment flow: request, accept, select, deliver, confirm. Atum launched recently, and for an early stage project, its documentation already establishes a strong technical foundation and a distinct identity. I'm not proposing that Atum simply publishes more content, I want to help translate that technical foundation into a reusable education and communication system. Think: - X threads and articles that make Atum's architecture easier to understand. - Explainers that help developers understand the payment flow and what they can build with it. - Research-backed articles and ecosystem communications that translate technical developments into meaningful stories. The goal isn't to dilute Atum's technical language. It's to help more people understand it without losing what makes the product different. I also noticed @PeteCooling's statement about harnessing your imagination if you want to work at Atum and it resonated with me. As a researcher, writer and content strategist, I'm constantly asking: How can I explain this more clearly? How can I make a technical concept easier to understand, remember and use? That's the kind of thinking I'd bring to Atum. Here's my proposal: I'd love to start with a small content pilot based on Atum's existing documentation. E.g one technical explainer and one X thread demonstrating how I'd approach product education. You can evaluate the quality of my work and how I think before deciding whether there's an opportunity for a long term collaboration. I've spent over three years working across Web3 research, technical communication and content strategy. Now, I'm looking for a team where I can put that experience to work. @PeteCooling, @atumhq I'd love to discuss this with you.

what’s that one thing you use AI for all the time? research? motion design? video editing? writing? for me it's research and when I open Claude or ChatGPT, I have to write a detailed prompt with context, instructions, preferences, examples, etc. I had to do this every single time, for every single research task. then I started learning about how to use AI better and I stumbled on the concept of workflows. I realised that even the project I created on Claude named "Web3 research" is not a workflow. you've only created a workflow when you've built a system that repeats itself anytime you feed it a task. building a workflow means that I can say "research on Avalanche" and the model knows the questions to ask, the type of answer I'm looking for, where to look and how to structure the final output without me describing what I want everytime. I have built 3 workflows for 3 different tasks in this month and here's how I did it. you don't need a crazy automation setup to a workflow. you can just follow 5 steps. 1. Pick that ONE task you repeat. e.g research for bounties. 2. Write down how you currently do it. e.g I usually research the problem, target audience, infrastructure, etc. this becomes your process. 3. Turn the process into instructions. tell AI “whenever I give you X, do A → B → C → D.” add your rules, preferences, source requirements, output format, examples, etc. 4. Give the workflow a clear input and output. for example: input: research this project. process: find primary sources → extract evidence → verify claims → organize findings. output: a structured research pack. now you've got something repeatable. 5. Test it. Break it. Fix it. run the workflow on a real task to see where the AI gets confused, skips steps or gives you rubbish and update the instructions. turn it into a reusable prompt, custom GPT/project, skill, structured workflow, or eventually an actual automation. and that's it. if you're doing the same thing every day and giving AI the same instructions every day, you might not need another prompt. you might need a workflow.

back to what we came into the room to do. good morning guyss how are you doing?