If you’ve seen all the previous Lenovo CES 2026 leaks by Windows Latest, you might’ve noticed some kind of AI in all of them. Well, the truth is that there is more, and our sources say that the company has a whole section of proofs of concepts at CES 2026, and we have got exclusive information and pictures of all of them.

Lenovo’s booth at CES on January 6 in Las Vegas will have a few personal AI computing concepts, starting with the new ThinkPad Rollable XD Concept, which we already leaked, the Lenovo AI Glasses Concept, a Personal AI Hub Concept, and a Smart Sense Display Concept.

Of course, there would be other proofs of concepts as well, like the Lenovo Legion Pro Rollable, and some Motorola concepts. But those four devices have more AI baked into them, says our sources.

While the ThinkPad Rollable remains to be the out-of-the-world, almost sci-fi-like device, the Lenovo AI Glasses Concept is also intriguing, especially since it looks like regular glasses, unlike the fully covered Apple Vision Pro or the Samsung Galaxy XR headset.

Lenovo AI Glass Concept for CES 2026

Based on the promotional image we received, the Lenovo AI Glasses Concept seems to be the company’s attempt to make an everyday wearable personal AI headset. Unlike bulky mixed-reality devices, these glasses look and feel close to regular eyewear, but include AI, multimedia control, and multi-device connection.

Our sources describe the AI Glasses Concept as a lightweight, always-connected companion that works wirelessly with a paired smartphone or PC. You can interact with the glasses using touch and voice controls built into the frame, with features that enable the AI Glasses to handle calls, control music, or access information without pulling out another device.

We are told that Lenovo is also positioning these glasses as a productivity aid, with features like a built-in teleprompter for presentations or speaking engagements, which could be useful for professionals who spend time presenting or livestreaming.

The AI side of the experience is powered through Lenovo Qira, which may be the company’s yet-to-be-announced AI Assistant. However, it taps into the compute of the paired smartphones or computers rather than running everything on the glasses themselves. Our sources mentioned a disclaimer saying the connected device needs to have Lenovo Qira.

From what we were told, this enables features like sub-millisecond real-time live translation and image recognition (with a camera on the front of the frame), giving users contextual information about what they’re seeing or hearing. There’s also a “Catch Me Up” style overview that shows summaries of notifications across multiple devices at the start of the day.

Comfort is clearly a priority too, with the Lenovo AI Glasses weighing around 45 grams and providing up to eight hours of combined productivity and entertainment on a single charge.

Note that our sources mention that this figure is based on Lenovo’s internal testing during November 2025, with 50% display brightness, and so the results may vary for different users.

Key features of the Lenovo AI Glasses Concept (as shared with us):

• Lightweight design (~45g)
• Wireless tethering to phone or PC
• Touch and voice controls
• Hands-free calling and music playback
• Built-in teleprompter mode
• Live translation support
• AI-powered image recognition
• Cross-device notification summaries
• Up to 8 hours of battery

Lenovo Personal AI Hub Concept (Project Kubit) for CES 2026

Based on information shared with Windows Latest, Lenovo’s Personal AI Hub Concept, internally referred to as Project Kubit, is a personal edge cloud device.

Unlike other AI companion devices, which need the cloud for compute, this concept is built as a high-performance personal AI system meant to support AI-enabled applications across PCs, smartphones, wearables, and smart-home solutions.

What makes Project Kubit powerful is its internal hardware. Our sources say the Personal AI Hub houses two ThinkStation PGX compact AI workstations, both built on the NVIDIA GB10 Grace Blackwell Superchip. These are behind a transparent touch-screen display, which is the primary interface for interacting with the system.

The ThinkStation PGX itself was announced recently as Lenovo’s first workstation built specifically for AI development, and it’s designed to run locally, without relying on cloud compute.

Each ThinkStation PGX is powered by NVIDIA’s GB10 Grace Blackwell Superchip, providing up to 1 petaflop of AI performance and supporting AI models as large as 200 billion parameters, or up to 405 billion parameters when two PGX systems are linked together.

The Personal AI Hub Concept, based on our sources, uses the same dual-PGX configuration for scaling AI workloads. If it has the same specifications, then, using the 128GB of unified LPDDR5x memory per system, NVIDIA DGX OS, and the full NVIDIA AI software stack preloaded, the hub may be capable of running and inferencing large models locally.

Instead of sending private data from PCs, phones, wearables, and smart-home devices to external servers, the AI processing happens inside the Personal AI Hub Concept itself, good for the privacy-conscious out there.

Our sources described this as a way to keep personal AI workloads local, and connected to the user’s own ecosystem. The transparent touchscreen and touch-and-voice interface sit on top of what is essentially a clustered AI workstation.

We believe that Lenovo’s idea with the Personal AI Hub Concept is to bring workstation-class AI compute into a personal, desk-friendly form factor, instead of depending on remote cloud infrastructure.

According to the information we received, the hub continuously harvests and aggregates data across platforms, giving it the ability to perform deeper analytics and AI-enabled applications that evolve over time based on real usage patterns and interactions that the user makes.

Our sources describe the experience as a powerful, always-available personal AI layer that draws data from the interactions between an individual’s devices every day.

Key features of the Lenovo Personal AI Hub Concept (as shared with us):

• Internal project name: Project Kubit
• Personal edge cloud AI device
• Supports PCs, smartphones, wearables, and smart-home devices
• Dual ThinkStation PGX compact AI workstations
• Powered by NVIDIA GB10 Grace Blackwell Superchip
• Transparent touch-screen display
• Touch and voice interaction
• Cross-platform data aggregation
• High-performance personal AI computing device

AI-Powered Display Concepts

Alongside its wearable and personal AI computing concepts, Lenovo is also preparing a set of AI-driven display concepts for CES 2026.

Based on the information we received, the two concepts shown are the Smart Sense Display Concept and the AI-Powered Personalized Display Concept. While they target slightly different use cases, both are built around the idea that displays should understand users, devices, and environments in real time.

Lenovo Smart Sense Display Concept

From what our sources describe, the Lenovo Smart Sense Display Concept is designed to act as a central, intelligent hub for multiple personal devices. In addition to traditional wired connections, the 27-inch UHD display can automatically connect wirelessly to multiple devices, including a smartphone, laptop, and tablet.

Users can cast content from their devices and work across multiple screens simultaneously. The display also supports natural-language interaction with a built-in smart assistant, allowing users to access devices or applications using touch or voice commands.

The Smart Sense Display Concept supports flexible layouts and drag-and-drop interactions, making it easier to move content between devices and organize workflows.

Key features of the Lenovo Smart Sense Display Concept:

• 27-inch UHD display
• Wired and wireless multi-device connectivity
• Auto-connection to phone, laptop, and tablet
• Wireless content casting
• Natural-language assistant via touch or voice
• Flexible window management
• Drag-and-drop multitasking across devices

AI-Powered Personalized Display Concept

The second Display concept from Lenovo shifts attention to comfort and well-being. According to our sources, the AI-Powered Personalized Display Concept is a multi-user display that adapts to who is using it and under what conditions.

This display concept can adjust brightness, color temperature, and visual settings in real time based on user profiles, time of day, and geographic location. These adjustments are meant to support circadian rhythms and reduce strain during long usage sessions.

The concept also includes real-time wellness monitoring, with the display responding to factors such as posture, eye fatigue, and viewing behavior.

Key features of the AI-Powered Personalized Display Concept:

• Multi-user display profiles
• Automatic circadian-based adjustments
• Personalized brightness and color temperature
• Real-time posture awareness
• Eye-fatigue monitoring
• Context-aware visual adaptation

As all these are concepts, we have no idea when or if they’ll come to the market, but that doesn’t take away from the possibilities that these offer. There are a bunch of people and businesses who despise the potential privacy concerns of AI, but need it nonetheless, and for them, the Personal AI Hub makes a lot of sense.

Add to the fact that it can even connect to the Lenovo AI Glass Concept, which in turn can connect to your smartphone, PC, and tablet, calls for a truly AI-powered ecosystem. Not to mention the regular glasses look of the AI Glass Concept can potentially make it a competitor to Meta’s AI glasses and even the Apple Vision Pro, despite it being in a completely different segment.

The post Exclusive: Lenovo Smart Glasses connect to PCs and phones, may challenge Apple and Meta appeared first on Windows Latest

how we put ai on like another jacket, threw a course on top, and ended up with burnout 2.0

---

for the last two years you’ve been sold the same dream on repeat:

“plug in ChatGPT / Cursor / a dozen prompts — and you’ll finally breathe.”

the punchline for a lot of people is the opposite:

• one more task system,
• one more chat,
• one more course on “how to prompt correctly,”
• and that sticky feeling that you’ve got less energy and more chaos.

“i’m doing everything right: watching webinars, saving ‘100 prompts for productivity,’ running tasks through ai… and by evening it still feels like i didn’t move anything forward.”

sound familiar?

in this piece we’re offering a different diagnosis:

often you’re not burned out. \n you’ve got context obesity.

and yeah — ai, courses, and “magic prompts” are basically tossing more firewood on the pile.

---

why you can trust this text

we’re the team behind AI Mindset and {context} lab. for the last two years we’ve been running labs and communities for people who build an ai stack and a personal operating system — not just “mess around with prompts.”

more than 700 people have gone through our formats already: product folks, developers, hr, consultants, founders.

from the latest lab:

• 60+ participants in a single cohort,
• 15 deep interviews about their experience,
• segmentation into four types of people (from “tech skeptics” to founders).

and they all share the same plot: people come in “for ai tools” — and hit a wall where their brain can’t handle the context configuration they’re living in.

this article is our attempt to describe that layer honestly.

---

classic burnout vs context obesity

classic burnout is about depletion. \n context obesity is about overloaded working memory.

if we simplify hard: your head has a limited “context window.” only a few items fit in there at once. when those items get closed — the system chills out. when they don’t — you drag tail-ends behind you.

psychology has been pointing at a couple key effects for a while:

• the zeigarnik effect — unfinished tasks stick in memory and hum louder than finished ones (read more);
• attention residue — part of your brain stays in the previous task even after you switch (sophie leroy’s research).

in a world where you have one project and one notebook, it’s manageable. but when you’ve got:

• a work project,
• a pet project,
• ai experiments,
• side education,
• several active chats with colleagues and communities,

those tails start eating your ram.

what burnout looks like

classic burnout:

• exhaustion, cynicism, lower performance;
• “treated” with vacation, job change, therapy;
• emotional + physical “i can’t do this anymore.”

context obesity:

• the feeling of a stuffed head and a “second job” maintaining systems;
• treated by rebuilding your context configuration, not by one more vacation;
• about the architecture of tasks, tools, and expectations — not just hours in the office.

these two can overlap. but a lot of people who self-diagnose burnout are actually living in chronic context overload.

---

what a day looks like with context obesity

let’s call them Alex, a product manager in a big company. but feel free to swap in yourself.

you’re not burned out, you’ve got context obesity

09:10 Alex opens the laptop and, on autopilot, boots up:

• corporate messenger,
• email,
• notion,
• ChatGPT.

somewhere in the background there’s a thought looping: “i should finish that experiment concept we came up with last week.”

10:30 on a call they discuss a new hypothesis. someone drops a link: “10 ways to use ai in analytics.” Alex opens it, scrolls, saves it to “read later.” meanwhile, a personal todo gets a new entry: “think about an ai assistant for reports.”

12:15 Alex decides to “speed up” and asks ChatGPT to draft an email to a client. the text is fine, but:

• facts need checking,
• it needs to be translated into company language,
• it needs to match the old thread.

so instead of 15 minutes for an email — it’s 25 minutes of review.

14:40 sprint planning time. Alex has:

• a backlog in jira,
• product ideas in personal obsidian,
• notes from the last retro in notion.

some tasks are duplicated, some contradict each other. Alex promises: “tonight i’ll consolidate everything into one system.”

18:30 Alex closes the laptop with this feeling:

“i was spinning all day, but not a single track feels finished.”

this isn’t about Alex “doing nothing.” it’s about nothing closing a loop: not the email, not the experiments, not the task system.

in interviews, our lab participants described it almost word for word:

• “my notes and notion turned into a second job.”
• “i open my system and realize i don’t understand my own system.”

and almost everyone showed up with: “i think i’m burned out.”

---

why “100 prompts for productivity” makes it worse

you’re not burned out, you’ve got context obesity

the industry responds to this state like this:

• “here are 50 prompts that’ll make you a super-manager,”
• “ChatGPT course: 3x your speed,”
• “5 tools that will replace your assistant.”

what happens when someone in Alex-mode walks into that content?

• every prompt list = another layer of expectations and guilt. now you’re not only behind on projects, you’re also “not using ai to the max.”
• every prompt creates new context. you generate dozens of drafts, ideas, lists — and all of them demand review and decisions. congrats, you’ve just opened new unfinished loops.
• ai doesn’t plug into a system — it plugs into chaos. if you don’t have clear architecture for tasks and meaning, ai simply increases the amount of text your brain has to chew.

people said it like this in interviews:

“i used ai every day, but it felt like another layer of work. the only thing that helped was when we started discussing the architecture of context — not just touching tools.”

so yeah: “100 prompts” isn’t medicine. it’s a new set of calories for an already overloaded context.

---

ai as a context-obesity amplifier

ai tools amplify the system they land in.

if you’ve got a clean process and sane load — they really do remove routine.

if your task configuration is already chaotic — they amplify the chaos.

you’re not burned out, you’ve got context obesity

there’s a useful term here: the “hidden tax of context switching.” when you jump between tasks and apps, your brain pays a reboot cost. across different reviews, estimates say it can eat up to 20–40% of work time.

some breakdowns here:

• https://www.basicops.com/blog/the-hidden-cost-of-context-switching
• https://reconfigured.io/blog/hidden-tax-of-context-switching-your-brain-pays-for-multitasking

add research on technostress — digital stress from too many tools and notifications. meta-reviews in recent years show that introducing digital tech without changing processes often increases stress and burnout risk at first, not lowers it (the review, and a popular retelling).

now let’s look at three layers of taxes in the “you + ai” bundle.

1/ integration tax

every new service means:

• figuring out how it works,
• fitting it into your current stack,
• designing how you won’t drown in its notifications.

in our interviews, many participants spent the first weeks not solving tasks, but maintaining the tool zoo.

2/ cognitive tax

ai writes — you review. the responsibility for meaning is still on you:

• check it didn’t hallucinate,
• adapt it to context,
• decide what to do with it.

that’s another stream clogging your working memory.

3/ the “the system will save me” illusion

especially nasty when ai lands on top of an existing graveyard: notion, obsidian, multiple todo apps.

instead of admitting “my system can’t hold reality,” we slap one more smart layer on top.

if the foundation is rotten, no ai penthouse will save it.

---

why vacation or a job change doesn’t fix it

you’re not burned out, you’ve got context obesity

when you say “i’m burned out,” you usually get advice like:

• take a vacation,
• change jobs,
• delete social media,
• “start from a clean slate.”

all of that can be fine, with one catch:

you bring the same context habits with you.

on vacation your brain keeps looping:

• unclosed promises,
• hanging projects,
• decisions you didn’t make.

a job change swaps the set of tasks — but doesn’t change how you:

• open too many fronts,
• keep everything in your head,
• are afraid to admit something isn’t important and let it go.

a new notion workspace without new logic turns into the same landfill in a month.

this isn’t willpower. \n this is missing an explicit system prompt for yourself.

---

humans need a system prompt too

large language models have a system prompt — a text that defines:

• who it is,
• what style it answers in,
• what it won’t do,
• what its default goals are.

you’re not burned out, you’ve got context obesity

most people don’t have that explicit layer. every morning we boot life in a mode like:

• “take everything that lands,”

• “don’t say no to tasks,”

• “reply immediately so you don’t feel guilty.”

  

\ at one lecture on context, a participant said:

“when i realized my head has a context window too, i stopped shoving everything in there.”

that’s basically the first sketch of a personal system prompt:

• what’s allowed for me right now — and what isn’t;
• what tasks don’t get to enter my ram at all;
• what rules i set for myself and for my ai assistants.

---

how to work with context obesity: three levels

spoiler: you don’t need to “burn it all down and go back to a paper notebook.”

you’re not burned out, you’ve got context obesity

you need to rebuild your context configuration honestly.

in our labs, three mandatory layers show up almost every time.

1/ capture: get the context out of your head

as long as your head is the only database, you’re doomed.

exercise: “15 unfinished loops”

• open a note or grab a piece of paper.
• for 7–10 minutes, write down everything that responds to “i should…”: “send the documents,” “talk to that colleague,” “figure out this service.”
• don’t sort or judge. the job is to offload.

people usually freak out at the volume. one participant said after doing it:

“i realized i’m not burned out — i’ve got a whole jira board open in my head.”

the simple fact these things moved from brain to text already brings relief.

2/ structure: give it a skeleton

after the offload, you don’t need to “optimize everything.” you need to format the chaos.

frameworks help here — not as religion, as a grid:

• the wheel of life — see where you’re objectively overdoing it (work, study) and where you’re crashing (health, relationships). breakdown, for example: https://quenza.com/blog/wheel-of-life-assessment/
• dilts levels — realize you’re fixing a skill when the problem is environment/values. description: https://www.skillsyouneed.com/lead/logical-levels.html
• yearly reviews like yearcompass — turn a chaotic year into a coherent story (site: https://yearcompass.com/).
• weekly cycles (for us it’s the arc method: appreciate → reflect → create) — end the week not with “i survived,” but with a set of concrete artifacts.

micro-format: burn / delegate / keep

once a week:

• open your list of unfinished loops.
• for each item, mark it honestly:
• burn — i won’t do it, and i’m brave enough to admit it.
• delegate — a person or ai can do it (drafts, data prep).
• keep — it’s truly important and belongs to the next few weeks.
• everything you “keep” goes into one system (any system) and stops living in your head.

often at this step the list shrinks 2–3x — and the guilt shrinks with it.

3/ your personal system prompt 2026

the last layer is turning all of this into explicit rules.

exercise: three rules for my brain-llm

imagine your brain is an llm and you can write its system prompt. answer in writing:

what three rules do i put into my system prompt for 2026?

it might look like this:

• don’t hold more than three active projects in my head. everything else goes to the backlog with an honest “not now” date.
• don’t plug in a new ai tool until i’ve lived in the current configuration for a full week. no new shiny toy until the old stuff is routine.
• every friday, close at least one loop: do / delegate / burn. so the week ends with completion, not hanging tabs.

then:

• run your current tasks through these three rules;
• see what falls off automatically, what needs to be handed off, and what actually stays yours.

for many lab participants this step alone changes their week configuration — without adding a single new service.

---

where the context lab came from

you’re not burned out, you’ve got context obesity

context lab was born from a painfully grounded question:

why do smart, motivated people with “good systems” in notion and obsidian still feel like Alex from this article?

we started gathering small groups and testing a hypothesis: if you give people language for context, offloading practices, and basic frameworks — does their “i’m not getting anything done” feeling change?

interviews show that this exact layer — understanding context and working with it regularly — is what people most often name as their main “click.”

context lab is a place where you don’t get another set of “secret prompts.” you build your context stack and your system prompt — and only then hang ai assistants and automations onto it.

second article in the series:

the productivity museum: why notion, ai, and task courses only make it harder to live

and in the third we’ll share a personal story of someone who drowned in obsidian, crawled out through their own system prompt, and along the way helped invent the context lab formats.

Ray Svitla \n stay evolving 🐌

As 2025 comes to a close, one thing feels clearer than ever: Excel is no longer just something you use. It’s something you belong to. 

This year brought major product innovations, many powered by AI, but it also delivered something just as meaningful: cultural moments that reminded us how deeply Excel is woven into work, learning, creativity, and even competition around the world. 

From celebrating a milestone birthday, to watching spreadsheets light up arenas and streaming platforms, to shipping some of our most ambitious product updates yet, 2025 was a year we’re incredibly proud of. And none of it would have happened without you. 

Let’s take a look back. 

A Cultural Year for Excel 

Excel Turns 40! 

In 2025, Excel celebrated its 40th birthday—four decades of helping people think, analyze, build, and decide more effectively. 

What began as a simple spreadsheet application in 1985 has evolved into a foundational tool used by hundreds of millions of people across industries, roles, and continents. Over the years, Excel has adapted to new technologies, new ways of working, and entirely new audiences, without losing the core flexibility that made it so powerful in the first place. 

We marked this milestone by reflecting on Excel’s past and, more importantly, its future: one where data literacy, accessibility, and creativity continue to expand. 

👉 Read more in Excel Turns 40: Join the Celebration!

 

The Excel World Championship Goes Mainstream 

If you needed proof that spreadsheets have officially entered pop culture, look no further than the Excel World Championship (EWC). 

In 2025, the competition reached new heights with larger audiences, more global participation, and unprecedented attention. What began as a niche idea has grown into a true esports-style event that proves how dynamic, fast-paced and thrilling Excel can be in expert hands. 

Watching competitors solve complex problems live under pressure and at speed was both entertaining and inspiring. It showed that Excel mastery is a real skill built through practice, creativity and deep understanding. 

👉 Read more in Congrats to the Winners of the 2025 MECC & MEWC!

 

Spreadsheet Champions Brings Excel to the Big Screen 

This year also saw the release of Spreadsheet Champions, a documentary that follows six students from different countries on their unique journeys to achieve excellence in competitive Excel. 

More than just a story about formulas and grids, the film is about community, curiosity, and the joy of solving problems together. It captured something we see every day across forums, classrooms, livestreams, and workplaces: Excel brings people together. 

For many of us on the Excel team, seeing these stories told so thoughtfully was deeply moving—and a powerful reminder of who we’re building for. 

👉 Read more in Celebrating the Premiere of “Spreadsheet Champions” at the Melbourne International Film Festival

A Breakthrough Year for the Product 

While Excel’s cultural presence grew, 2025 was also one of the most ambitious product years in recent memory. 

Agent Mode in Excel  

One of the biggest shifts came with Agent Mode in Excel—a new way to approach work that moves beyond asking for help, to delegating outcomes. 

Instead of manually building step-by-step solutions, users can now describe goals and let Excel reason through the steps: gathering data, applying transformations, and explaining results along the way. It’s a meaningful step toward making Excel not just reactive, but proactive. 

Agent Mode doesn’t replace expertise; it amplifies it. 

👉 Read more in Building Agent Mode in Excel

 

The COPILOT Function Arrives 

In 2025, Copilot became more deeply embedded directly into the Excel grid with the introduction of the COPILOT function. 

For the first time, users can call Copilot like a formula, bringing AI-powered reasoning directly into cells alongside traditional Excel functions. This bridges the gap between natural language requests and structured spreadsheet logic, unlocking entirely new workflows. 

It’s one of the clearest examples yet of how AI and spreadsheets can work together seamlessly. 

👉 Read more in Bring AI to your formulas with the COPILOT function in Excel

 

Formula Completion Gets Smarter 

Excel has always been about speed and precision, and in 2025 we made writing formulas easier than ever with improved formula completion. 

Smarter suggestions, better context awareness, and faster recommendations mean less time remembering syntax, and more time focusing on insights. Whether you’re learning Excel or pushing it to its limits, formula completion now meets you where you are. 

Small improvements like this matter. They add up to a smoother, more confident experience for everyone. 

👉 Read more in Introducing formula completion - A new way to write formulas in Excel using Copilot

 

Thank You for an Incredible Year 💚 

If there’s one theme that defines Excel in 2025, it’s this: progress powered by community. 

Every feature we shipped and every moment we celebrated was shaped by customer feedback, creator experimentation, MVP insight, and everyday use in the real world. You pushed us, inspired us, and reminded us why Excel continues to matter—40 years on. 

As we head into 2026, we’re excited to keep building with you. 

Thank you for being part of the Excel story. 

Unlock sales productivity with ChatGPT! Learn how to use AI for prospecting, outreach, discovery, and more. Boost efficiency and close more deals. Practical tips inside!

Unlock ChatGPT for marketing! Learn how to boost content, SEO, and campaigns. Discover real-world strategies to amplify creativity and drive results with AI. Optimized for search!

15 minutes to read

We’re about a year into the process of completely re-working Sdkbin to better support our needs here at Petabridge and, eventually, other third-party software vendors.

Sdkbin was originally built on an extremely flimsy CRUD architecture that violates most of my personal design preferences - you can read about the history behind that here. But to summarize, I tend to use the following heuristics when building software:

1. Prefer optionality-preserving designs - make sure your design decisions can be reversed or altered when things inevitably change.
2. Use as few moving parts as possible - most of Akka.NET is constructed this way.
3. No magic - if nothing magically works, then nothing magically breaks either.
4. Ensure that coupling happens only where it’s necessary - coupling usually needs to happen in your integration layer (i.e. your UI or HTTP API.) Your accounting system should probably not be coupled to your payments system.

Sdkbin’s original CRUD design violated all of these principles:

1. Used hard deletes, destroying data with no ability to recover or audit (impossible to reverse);
2. Relied heavily on AutoMapper-powered generic repositories (magic with lots of moving parts); and
3. Was highly coupled throughout - Stripe payment events served double-duty as invoices, for instance.

We’ve fixed a ton of these issues already, and one of the most important tools we’re using is Akka.Persistence. We’ve also made some significant improvements to Akka.Persistence in recent releases that made it much easier for us to accomplish our ambitious goals with Sdkbin.

Let me show you what we did.

Click here to read the full article.