Claude's Contribution to Streamlining Workflows

I turned the most annoying 20 minute task of my job into 20 seconds. Here's the step-by-step breakdown: For years, I've been drowning in business card requests. Emails at all hours, random texts, people stopping by my office - all asking for the same thing but never giving me complete information. I'd spend 15-20 minutes per request just coordinating between employees, our designer, and placing orders. It was literally the least important but most time-consuming part of my week. Working with Claude (Anthropic's AI), I built my first AI agent that completely automates this workflow. Now when someone needs business cards, they fill out a simple form in Teams. The moment they hit submit: ✅ Data automatically saves to SharePoint ✅ A ClickUp task gets created for our designer with all the details ✅ Task gets assigned with a 3-day deadline ✅ I get notified when it's ready for ordering What used to take me 15-20 minutes of back-and-forth now happens in seconds. Zero manual work on my end. The crazy part? This entire system was built in a few hours using tools we already had - Power Apps, SharePoint, Teams, and Power Automate. No coding required. Here's my biggest takeaway: AI agents aren't just for tech companies. They're for anyone tired of repetitive tasks eating away at time that should be spent on strategic work. I'm a marketing leader at an oil and gas services company, not a programmer. If I can build this, anyone can. What repetitive task is driving you crazy? Maybe it's time to automate it. Next up: I'm eyeing our expense reporting process 👀

Claude Code might be Anthropic's best product and their marketing team's biggest fumble. Multiple smart people in my network have been raving about it, so I finally decided to dive in. I'm pretty blown away. Yes, it's incredible for coding. But here's what surprised me: it's equally powerful for content workflows, data cleaning, and complex task sequencing. The tool is scary good at breaking down complex problems into subtasks and knowing exactly when to keep you in the loop versus when to just execute. But here's where I'm confused: The biggest barrier isn't the price. It's that you have to run everything through Terminal. For most people, Terminal feels intimidating, and they will never take the plunge. The funny thing is that it's actually super easy to pick up - you get comfortable in 1-2 hours (This is coming from someone who's not super technical) So why hasn't Anthropic built a more accessible interface? And why call it "Claude Code"? It's incredible at coding, but it's doing the product a disservice. This thing is more like "Claude Agent." I can't tell if I'm missing something obvious or if Anthropic's marketing team is fumbling a killer product with terrible positioning.

In a total of 1 hour and six minutes over the course of a few days I took a task that previously consumed 1-2 hours/week and reduced it to less than 15 minutes using Claude Pro. No other tools, no vibe coding, no stumbling around command line or hacking together Python code. Here was my process: Step 1: Sitting in the back of a Lyft on Friday night, I remembered I'd kept wanting to get to this task of automating my AI News Summaries. I create these for the team to help the agency stay in tune with the most important news of the week. I'd been slacking recently though, and what used to be a weekly updated was becoming a monthly update. I had all the logic I use in my head for which newsletters I review, what I look for, what is a priority, etc. I typed all of that out, but then arrived at my location and put a pin in it. Step 2: Cut to Monday morning. I found the thread in Claude and pulled the instructions it'd made into a new Claude Pro Project. Gmail integration took a couple of clicks. I generated the summary and reviewed it. I spotted some stuff I wanted to address, so I called those out and asked Claude to update the instructions. It did, and I generated a new summary for the past week. Step 3: I did my manual review of the past weeks' news and whenever I saw something I wanted to ensure was covered in the summary for the Seer team, I checked the generated summary. It was almost 100% in alignment with my inclusions, and I updated the project instructions one more time to improve alignment. Step 4: I reviewed the summary in full and removed some less priority updates that felt too noisy for the team's consumption. I thought about updating the instructions but for now I'd much prefer having too much news vs not enough. Step 5: I copied the text into my google doc repository that powers a Google LM Notebook. I am sure I could get Claude to write right to this doc but not a priority for now. I copied the update into our AI Slack Channel and came to LinkedIn to brag about my accomplishment 🤭 Really though, the lesson is this: the ceiling for what we can build is getting absurdly high. It's exciting, but there is still so much potential to build value through a couple of simple, easy to use tools and their native connections.