Hey PH

I d love to share a project I ve been building and get some honest feedback.

Circuit is a platform where posts aren t text, images, or videos they re interactive experiences.

Instead of explaining how something works, you publish the actual demo as the post so people can click, explore, and understand faster.

One of the coolest parts of my job is getting a front-row seat to how @marianaprazeres thinks about AI.
Memory feels table stakes in AI right now. But for @Meet-Ting, it s not just a log of the past - it s a living system that shapes how people schedule, work, and want to spend their week.
It s not just logistics - it s patterns around energy, priorities, and relationships over time.

Here are a few things we learned while designing and testing agent memory in production:

I m working on a new annotation and drawing experience in Phia that lets you draw and highlight while recording. My goal is to keep it fast, minimal, and natural, so annotations improve the video without feeling distracting.

I m sharing the current design and would really appreciate honest feedback. Does this feel intuitive to use while recording? Are the pen and highlighter styles useful? Is there anything missing, or anything you d simplify?

I m still iterating on this, and early feedback will directly shape what I build next. Thanks for taking the time to look and share your thoughts.

Hi guys,

Hey everyone

Thank you again for the support on Alpie Core, and the feedback from this community meant a lot to us.

Since then, we have finally released Alpie, our most advanced product yet. A full AI workspace where you can now see Alpie Core working in real workflows, and not just isolated prompts. You can use the model with files and PDFs, run research, collaborate with others in shared chats, and keep long-running context organised.

If you were curious how Alpie Core performs beyond single queries, this is where you can try it hands-on.

Hello!

After 7 years slogging away in management consulting and private equity, I;ve decided to slog away in the world of startups. I have a passion for sustaining small businesses and supporting successful transitions through the M&A transaction process. However, the system has failed the lower middle market and main street business owners. I built MainSquare to fix that. We take a value creation approach to making companies more valuable AND more sellable - organically. Business exits must start years in advance of a deal. We've built a platform that assesses business and generates value creation roadmaps instantly, then aggregates skilled partners based on the business' most critical needs.

With today s tools, translation (UI, copy, even video) is no longer the hard part.

What slows us down instead are things like tax, legal compliance, hiring, support, payments sometimes even geopolitics. The moment users show up from a new country, a product problem turns into an operating one.

2025 wasn't just another year. It was the year LLMs grew eyes, ears, and hands.
We went from models that could write a decent email to multi-modal systems that can reason, edit images, see videos, and - most importantly - execute actions.
For us at MindPal, this was the moment we had been waiting for since Day 1.
2 years ago, we started with a vision: that someday, AI would process multi-modal inputs, manage complex context, and use tools just like a human. Back then, the tech wasn't ready. We spent 2024 helping you with "prompt chaining" - squeezing out a slightly better blog post or summarizing a document. It was useful, but it was just assisting.
That era is over.
In 2025, we stopped just chaining prompts and started orchestrating intelligence. We turned MindPal into the multi-agent engine we always envisioned:
- Advanced Patterns: We moved beyond linear chains to Orchestrators, Loops, Evaluators, and Routing.
- MCP Support: Agents can finally talk to your tools - Google Drive, Notion, Dropbox - natively.
- Canvas: Agents aren't stuck in the chat anymore. They are building custom websites, slides, and sales collateral in real-time.
- Chat-to-Build: You don't have to figure out the logic alone. You can now chat back and forth with the AI at the very start to architect, refine, and improve your agents before they ever run.
- Deep Integration: With Custom User IDs and Session Context, your embedded agents aren't isolated. You can sync your own users' data from your system directly into MindPal embeds.
Was every bet a winner? No.
We tried to force computer use and browser use capabilities too early. We realized it wasn't robust enough for the way you work, so we cut the noise and refocused on reliability.
But the things that did work? The results were insane.
We are seeing users automate their entire sales intelligence stacks - from qualification to generating custom proposals. We are seeing agencies automate client deliverables, where their whole operation depends on MindPal to run. We are seeing industry-specific audits that require deep, advanced knowledge and reasoning being executed autonomously.
MindPal doesn't exist to help you save costs. We exist to help you scale your business to a level that wasn't thinkable before, by scaling intelligence.
In 2026, we are taking a fresh take on AI automation.
Too many solutions right now are just "Software 2.0." They are obsessed with connecting App A to App B, with an AI sprinkled in between. It's rigid.
We believe the future belongs to Agency.
We are moving beyond just automating rules one after another. We are building the platform that leverages an agent's ability to think, reason, and wield tools with flexibility. We want you to scale work that was previously impossible, at a scale that was previously unimaginable, not just replicate what you were already doing.
Our goal is to be the number one AI agent orchestration layer for business.
2026 is going to be incredible. LFG