What if your PC could actually listen, understand, and do the work for you?

I’m building that idea into something real.

Meet AURA — my personal AI voice assistant for Windows.

Instead of building another chatbot that only answers questions, I’m experimenting with an AI agent that can actually interact with my computer and help with everyday work.

So far, AURA can:

• Have live voice conversations • Convert speech to text and respond with voice • Open applications like Excel, Power BI, Chrome, and Windows Settings • Search the web and YouTube through voice • Perform basic PC and file operations • Understand requests through Gemini • Execute actions through a safety-controlled automation layer • Continue listening after completing a task

But this is only the beginning.

My long-term goal is to make AURA understand my actual workspace — including what’s happening on my screen and inside applications — so I can simply say:

“Open my Excel file, check the sales data, clean the date column, and add a formula for total revenue.”

And let AURA handle the workflow.

I also want it to become useful for my own Data Analytics work — from Excel and Power BI workflows to repetitive desktop tasks, research, and automation.

I’m building AURA with Python, AI, voice technologies, desktop automation, Excel, Power BI, and LLMs, while keeping safety and human control at the core.

It’s still far from the final vision.

But that’s exactly what makes the project interesting.

I’m not just learning how AI agents work. I’m trying to build one that I can actually use every day.

Building in public. Learning by building. 🚀

#AI #AIAgents #Python #AIAssistant #VoiceAI #DesktopAutomation #DataAnalytics #PowerBI #Excel #Automation #BuildInPublic

The screen-reading step is where I would put the guardrail. Once AURA can read what is in the window, everything on that screen is input - and a page it opened from a voice search can hold a sentence written to be read as an instruction. I spent a while trying to rob my own...

😂 Honestly, I didn’t set out to build a public product.

I built ALAN OS for a very practical reason: my own design team was spending too much time moving information between tools, checking whether a render still matched the floor plan, and rebuilding the same decisions for every new view. I wanted one place where the project could stay coherent from the first conversation to the final handoff.

The first version is now a working, end-to-end spatial design workbench.

Here is the flow:

• Project brief — capture who will use the space, how they live, what matters to them, the city, the area and the real budget before talking about style. • Plan analysis — turn a floor plan into a clear set of walls, openings and functional zones, with a confidence check and a human confirmation step for uncertain dimensions. • Spatial model — keep one shared model as the source of truth, so renders, 360° views and presentation pages are based on the same geometry. • Style master — record the project’s palette, materials, lighting, camera language and furniture identity, then lock that “style DNA” before batch rendering. • Render set — bind camera points to the model and generate a consistent set of views instead of a collection of unrelated images. • 360° panorama — move from a fixed camera into a roamable scene while keeping the same doors, views, materials and furniture positions. • Presentation & handoff — turn the confirmed design decisions into a cleaner package that a client or team can actually review and use. • Cost estimator — calculate design and construction fees from area and unit rates, with room for discounts, furnishings, allowances and tax when needed.

The task queue and consistency gate are there to make the process visible. You can see what is running, what is locked and what is waiting. If the geometry, style master or camera coordinates drift, the batch is meant to pause instead of quietly producing a beautiful but incorrect image.

The cost page is already working in USD. For this 286 m² example:

Design: $200/m² → $57,200 Construction: $2,000/m² → $572,000 Total: $629,200 Blended rate: $2,200/m²

This is still v1, but it is already capable of supporting the full workflow. It is not a magic “make me a house” button, and it is not meant to replace design judgment. It is a way to keep the judgment, the geometry, the images and the budget connected so the team can spend more time designing and less time repairing handoffs.

The next layer I want to add is the part that makes a project truly ready for delivery: a detailed soft-furnishing schedule, lighting plans, water and electrical (MEP) layouts, and a clearer roll-up of those items into the total project price. I also want the calculator to handle currency exchange rates and tax rules, so a project can be priced consistently across countries instead of being rebuilt in a spreadsheet every time.

I built it for my own studio. The unexpected part is that other people started liking the idea too. That has me thinking about whether ALAN OS should eventually be open source. I haven’t made that decision yet.

If you have a good idea, a workflow pain point, or a feature you would genuinely use, please leave a comment or send me a message. I’d really like to hear your thoughts and build the next version with the people who will use it.

I like how ALAN OS treats the spatial model as the source of truth instead of another isolated render tool. The handoff from brief to style DNA to consistent views feels like a genuinely thoughtful systems workflow.