Heartspace.AI

Today I have been able to reroute most of Heartspace.AIs model router to be able to make the most use of being part of Google for Startups and to be able to let our users choose model provider themselves (or let our router do it for them). It’s funny how many moving parts includes. And how different things like streaming responses and thread handling works when working with different APIs. All of that are engineering problems we have had to solve once and now I resolving them but in a more streamlined and modular way. What will this lead to? 1. Users willl be able to choose their favorite model in Heartspace. If you love Claude sonnet 3,5 or any of the Gemini or OpenAI models you will be able to choose them by yourself instead of our model router choosing for you. 2. We will be defaulting more toward Gemini since Googles models are in general outperforming other models when it comes to content creation and because we can offer so much more from our inference approach when we can leverage the free credits Google has given us. Very generous program by the way! If you are an startup check it out! Google is doing awesome stuff for the builder ecosystem!This allows us to increase output quality and consistency without increasing prices. Most of the day I had the shit kicked out of me by resilient front-end bugs (front-end is still not my favorite thing) but progress is being made and by the end of the day I had a early version up and running. Still some things left to do but it is so good so can’t wait to release this to our users! Another cool thing is that we also got accepted into one of Cloudflares programs today and will get free hosting for a least a year from today. Saves us up to $5 000. Super grateful for that 🙌🏼🙏🏼.

Platform Updates: Enhanced UI, Editor & AI Co-Writer! ✨ We've been working on some updates to make using the platform more delightful: We've updated the look of some buttons, and improved the tooltips so they're more helpful. Links in the editor are now clickable, text flows better when using the co-writer, and formatting should behave more predictably. We've also cleaned up some extra line breaks (we still get them sometimes but working on a full fix for this!) You can now stop AI text generation if you need to, it delivers content faster, and you can regenerate the AI's output. We've also improved how it handles formatting and where it puts the text. Also the user experience should be overall smoother and to put it in clear terms: it looks a lot nicer 😊. Our goal is to make the platform more responsive and user-friendly, so you can focus on creating great content. Still Working On: Better formatting with lists from the AI. Smoother content streaming into the editor. More to come! https://lnkd.in/gWDr_Mmk

Heartspace.AI now has a public roadmap! For 2025, we're taking "building in public" seriously. What our users ask for becomes what we build next. We want to be 100% open and transparent, sharing both what we're doing well and where we need to improve. We know that we have a lot of both! If you're interested, check out the initial roadmap here (it's still a work in progress): https://lnkd.in/ec74HVH2

I asked LinkedIns learnings AI-assistant about how to battle self doubt and second guessing decisions (which I often find myself doing for no reason whatsoever). Yeah, you might not know that but LinkedIn Learning has its on internal AI-assistant. It can be super helpful when it comes to professional challenges. Like whenever you feel stuck and feel you need some knowledge or motivation and want to avoid the rabbit holes of social media. You know the feeling when you start searching for something on Youtube and end up with something entirely different. This one is confined to the library of LinkedIn learning (mostly) and acts like your personal coach and can suggest courses that fit the challenges you are meeting. For my challenge on second guessing my decisions it recommended this course. It was a really great course full of micro learnings and important reflections on how to stay aligned with goals and commitments and I felt like I got real tangible and concrete tips on how to stop the energy loss of fighting myself whenever I start second guessing my decisions and priorities. The analogy to swimming and recognizing the “tides” was really on point and made me realize how I sometimes let what is happening externally on a macro-level question too much of what I in reality already have put a lot of effort and time into making sure to validate. In those times I want to become better at recognizing those patterns and seeing when something really is “a change of scenery”, a “signal” or something important or when it’s just “noise” that diverts from focusing on what’s really important. I can really recommend both the course itself and LinkedIn Learnings AI-assistant. Super good stuff!

What do you think? Will ”agent-cies” really replace agencies? If so: what do you think will be the key deciders whether a company will be open to considering AI-drive ”agent-cies” over the traditional alternatives they have? Interesting reflections by Greg Isenberg! Worth giving a thought or two!

My biggest promise to myself this year isn't about results. It's about showing up. About committing. About consistency. Because if I've learned anything from trying to build a startup like Heartspace.AI, it's that outcome-based goals, while important and inspiring as a north star, are insufficient on their own. They need to be paired with consistency-based actions. What actions will I take, what will I feel and think? What state will I put myself in? Where and how will I show up? And what will I accomplish? So in 2025, I've decided to be kinder to myself regarding outcomes. Outcomes aren't always within our control. They depend on more than just our own efforts, and frankly, it's a little egotistical to believe we have complete control over everything. But what I do control is how I respond to those outcomes. How I regulate my mood and emotions. How I set my intentions. And ultimately, how I choose to spend my time thinking, speaking, and acting. We have an ancient Persian saying, central to the Zoroastrian tradition: "Pendare nik, goftare nik, kerdare nik." This aligns strongly with Buddhist principles of wholesome intentions, speech, and action. It focuses on the foundation of our intentions, thoughts, words, and actions, and what motivates them. If we truly want to see our long-term impact on the world, examining our daily consistency in aligning with what we truly believe in is a good indicator of where we'll be at the end of the year. I like how Github gamifies consistency through its commit map. Every time you push code to Github, your profile gets a little greener. I've adopted this approach for other areas of my life. In my office, I now have a "commitment map" for my 2025 goals. It serves as a daily visual reminder: have I committed to what I intended? 2025 has just begun, and I've already committed code 54 times on Github in seven days. By the year's end, my aim is to share a fully green table with you, and a fully green table for my other goals: Laugh and joke with friends and family every day. Talk to at least five current and potential users/customers every day. Improve one thing in the product based on customer feedback every day. Engage in physical activity until I get warm and need to catch my breath every day. Share my journey and learning on LinkedIn, with you guys, every 2-3 days in the week. Especially when (not if) I fail at what I have committed too. Or when things do not turn out the way I have hoped or thought. And what I have learnt from those twists and turns. As they happen. Because I know that will help me learn and stay committed. These are my commitments for 2025. This is my "take" on resolutions and new year's goals. It might not be for everyone, but for me, it feels more inspiring and attainable than anything else. Even on challenging days, I know I can do some of these things. And even if I occasionally stumble, committing and trying to show up as I intend makes me feel excited and ready for 2025.

Proaktiv redovisningsbyrå sökes – byrån som förstår det lilla startupbolagets verklighet och tar ett helhetsansvar, finns ni? Hej LinkedIn-nätverk, Jag är på jakt efter en ansvarstagande redovisningsbyrå som kan hjälpa mig att fokusera på att bygga Heartspace.AI och samtidigt ha full koll på siffrorna utan att behöva göra byråns jobb. Några saker som är viktiga för mig: • kvittohantering ska skötas så smidigt det går för oss som företag (dvs troligen lite mindre smidigt för byrån) så jag slipper manuell hantering och repetiv kontering av varje kvitto. Enbart nya utgifter behöver ju kollas av. Det ska ske på enklast möjliga sätt. Om kvittot landar i en delad mejlkorg så ska vi kunna slippa ladda upp det i bokföringsystemet och kontera det själva varje månad. Vill inte sitta och ”boka” samma kvitto varje månad och samtidigt betala för en s.k. ”digital” redovisningstjänst. Vill ha en byrå som tar ansvar som på den gamla goda tiden. • Viktigt att byrån tar helhetsansvar och är proaktiva. • Månatliga avstämningar med tydliga rapporter och koll på aktiveringar/eget kapital är jätteviktigt. Vi är ett SaaS-bolag med månadmoms och där siffrorna är viktiga styrmedel och aktuella nyckeltal avgörande för hur vi kan expandera vår verksamhet. Så det får inte släpa efter med rapporter. Det måste skötas prickfritt. • Senior konsult som följer upp och ger insikter, så vi som grundare kan vara mer ”hands off” och fokusera på att bygga bolaget men ändå välinformerade och insatta i ekonomin. Vi siktar på att teckna ett nytt avtal med start från 1 februari och vill gärna ha relevanta offerter i veckorna. Har du några bra tips eller känner någon som kan passa? Tagga gärna eller hör av dig i kommentarsfältet! Vi kan betala skäligt/marknadsmässigt för ovanstående hjälp. Vi vill ha kvalitet och professionalitet och det är viktigare för oss än pris. Sen kommer vi förstås jämföra offerter men priset är inte den mest avgörande faktorn. Tack på förhand 🙏🏼. /Sohrab

Curious about how we handle links in Heartspace? It's simple! Click on any text in your feed, and it directs you to the original article. This opens in a new tab or window, depending on your browser. When you insert links into the editor, just press space to make them clickable. No need to right-click; a simple click will do! We can also merge multiple texts into one summary. The links will remain clickable and integrated within the text. This means you can easily reference sources in your texts and give proper credits in your content! See the video for details and more in this helpcenter post: https://lnkd.in/exC9dA2f

Kickstarting the year with a new place to work from. Heartspace.AI is moving into SQUARE. A small room but hopefully we can grow from here. Come say hi if you are in Liljeholmen or around. The year has just started and I am excited to see what it will bring. Wishing you all a great start of 2025! 🙌🏼🙌🏼🙌🏼

This is big. CAD designs are the basis for a lot of industries and most of the ways we make any physical products that you use in your daily life. They are also everywhere in all hardware, constructions and production. Industrial design take a lot of time and there is a constant need for more people who can turn ideas into designs and design into code that machines can turn into products. Now that process can go from months and weeks to minutes. And in the near future if you have an idea for a new product or solution that is in the hardware space you will be able to iterate on it just like we can iterate today on software products. Imagine what this will do for the rate of innovation and how many more solutions that actually get somewhere before they are late to market. But will this not make everyone in industrial design unemployed? Nope. On the contrary more jobs will be created in the industrial design space then ever before. But different jobs. This will do for industrial designers what Cursor has done for programmers and Heartspace.AI does for content creators. Enable them to focus on deeper and more complex an strategic design tasks which will enable them to iterate faster and create better products. The roles change and people become directors and curators and editors instead of being slowed down by friction in their work. This will propel humanity forward. A lot. Video credit: @j_d_noone on Twitter/X.