Activity
Mon
Wed
Fri
Sun
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
What is this?
Less
More

Memberships

AI Automation Society

418.9k members • Free

🇹🇭 Bangkok IRL

186 members • Free

13 contributions to AI Automation Society
A tiny trust feature I added for client work
Built a small thing this week that I'm weirdly happy with. When I'm working inside a client's system, a little icon quietly appears in the corner of their dashboard. It's not there any other time. When I'm actually in, it glows, and the second I'm out, it disappears. If they hover it, it tells them plainly what's happening: maintenance is being done right now, by us. The reason I like it: one of the quiet worries clients have with this kind of work is "how do I even know when someone's in my system?" Instead of making them ask, the system just tells them, in a way that doesn't interrupt whatever they're doing. Presence you can see, that turns itself off when it's not relevant. Cost almost nothing to build and it's done more for trust than anything I've said on a sales call. Anyone else build these little transparency touches into what you hand clients?
Framework to spot a real problem to solve
**❌ I wasted 40 hours on this automation** Built it perfectly. Clean code. Worked great. Nobody wanted it. Why? I didn't validate first. I ASSUMED it was a problem. Turns out: - Only 1 person had it - It took 30 seconds to fix manually - They weren't willing to pay That was a $0 automation. Then I learned: Ask 5 people FIRST. **The questions I ask now:** 1. "How often does this happen?" 2. "What does it cost you?" (time or money) 3. "Would you pay to fix it?" 4. "Who else has this problem?" Get YES on all 4 → I build it Get NO on any → I skip it **Have you ever built something nobody wanted? Share in comments.** Honestly, I think it's the fastest way to learn 😅 and some more examples are there which I am added into the pdf in the post , you can check and build them!! questions: - 1. what your learning to building ?? 2. any addon? 3. how much helpful it can be for you!?
Framework to spot a real problem to solve
0 likes • 9h
This is the lesson that costs everyone their first 40 hours, glad you're passing it on. The one thing I'd add to your four questions is a fifth: are they already paying to solve it some other way. A clunky manual workaround, a person they've hired, a tool they tolerate, that's the strongest signal there is, because it means money is already moving toward the problem. When someone says they'd pay but nothing is currently bleeding, they usually won't. Existing spend beats stated intent every time.
Replacing Power BI Dashboards with AI-Coded Dashboards: Real Saving or Hidden Cost?
I’ve been seeing people replacing Power BI dashboards with dashboards built using AI coding tools like Claude Code. Some influencers are also selling this idea as a way to reduce license costs and build dashboards faster. Has anyone here tested this or actually implemented it in a real business workflow? Did it really save money? How was the maintenance, governance, sharing, security, and user adoption compared to Power BI? I’m curious to understand where this approach makes sense and where Power BI still wins.
1 like • 9h
Good question to ask before you switch, because the license savings are the easy part and usually not where the real cost is. I've gone code first for a lot of internal stuff and the honest tradeoff looks like this. AI-coded dashboards win when it's yours to own, you want it in git, and you need something custom that a BI tool fights you on. Power BI still wins the second other people need to build, share, and maintain their own views without calling you. The hidden cost isn't building it, it's who maintains it a year from now and whether a non-technical team can actually touch it. If the honest answer is "only me," code is great. If a whole department leans on it, that's where Power BI earns its license.
How to step up a second AI brain?
I've been using Claude Pro for a while and I want to build an "AI second brain" with it at the core. A few specific problems I'm trying to solve: - Remembering what I've already learned or researched - Speeding up how I consume and retain information - Getting help with writing and content — not just one-off prompts - Staying on top of tasks and projects without losing context I'm leaning toward combining Obsidian with Claude, and possibly using Claude Code to build automations around it (things like scripts to summarize notes, surface past research, or sync context). But I'm not sure how to connect these pieces in a way that actually works day-to-day. For anyone who has built something like this: → Where did you start — notes, tasks, writing, or something else? → Has anyone integrated Obsidian with Claude to auto-summarize notes or pull up past research? How did you set that up? → What's one thing you wish someone had told you before you started? Looking for real-world experience, including the parts that don't usually get talked about.
0 likes • 9h
You're on the right track with Obsidian plus Claude Code, that's basically what I run. A few things I'd pass on from doing it a while. Keep the brain as plain markdown files, not locked inside some app, so any tool can read and write it and you're never trapped later. Don't try to automate retention up front, the retention comes from the notes being good and searchable, the automations are just there to surface the right note at the right moment. And give Claude Code a file that tells it how your brain is organized, otherwise it re-learns your structure every session. Start with one loop that genuinely saves you time, like summarizing and filing new notes, and grow from there instead of building the whole thing at once.
The Next AI Advantage Isn't a Better Model Everyone is chasing the next model.
Bigger context. Faster responses. More reasoning. I think they're looking in the wrong direction. The next competitive advantage won't be the AI you use. It will be the memory you build around it. Imagine two people using the exact same model. One starts every conversation from zero. The other has years of documented decisions, workflows, lessons learned, failed experiments, writing style, business rules, and personal frameworks. Same AI. Completely different results. We're moving from prompt engineering to memory engineering. The people who systematically capture and organize their thinking today may have an enormous advantage in a few years—not because the AI is smarter, but because it understands their way of thinking. Maybe the most valuable asset we're building isn't another prompt library. Maybe it's a second brain that grows with us. Question: If your AI could permanently remember only one thing about how you work, what would you want it to remember? #AI Email Agent (9/20/24) #RAG Chatbot AI Agent (9/22/24) @Nate Herk
The Next AI Advantage Isn't a Better Model  Everyone is chasing the next model.
1 like • 9h
You're hitting on the thing I've gotten most sure of. The model is basically a commodity now, everyone can rent the same one. What nobody can copy is years of your own decisions and rules and dead ends written down somewhere the AI can actually reach. I run most of my company this way now, and the value was never the model, it's that the model wakes up already knowing how we work. Same tool, totally different output, and the difference is entirely what you built around it. Memory engineering is a good way to put it.
1-10 of 13
@jay-mansperger-8068
Founder of my silent partner holding company.

Active 6h ago
Joined Jun 11, 2026
Powered by