Shortly after we launched portfolios in July, learners started asking for a follow feature. They wanted to know when specific people in the community were building something new.

It seems like a fairly ordinary product request, there’s something much bigger behind it. Following has always existed in service of content. You subscribe to someone on YouTube because you want to watch their next video. You follow someone on X because you enjoy their writing. Instagram grew around photos, TikTok around short-form video. Every major social network has been built around a particular type of content, with the social graph emerging around the people who consistently create it.

Portfolios have never worked that way. They’ve traditionally been static records of work - something you put together when you’re looking for a job, fill with your best projects, and then mostly forget about until the next time you need it.

But that changes when a portfolio is created simply by building. Instead of a carefully curated snapshot, it becomes a running history of someone’s work. You can see what they’re building, how often they build, what they’re interested in, and how their skills are evolving over time.

With skills changing so quickly now, NextWork portfolios and the ability to follow other builders removes a middle layer that historically slowed down the transfer of knowledge. The portfolio is no longer a destination a recruiter visits once while making a hiring decision. Instead, it’s something people want to return to because it’s the best place to get the latest and greatest from people in the field. Following their portfolio gives you a direct view into that work as it happens, without requiring them to turn it into a tutorial, a LinkedIn post, or a YouTube video first.

I think that creates the beginnings of a different kind of creator economy. The people who build large audiences won’t necessarily be the best writers or video creators. They might simply be the people who consistently build interesting things.

And once those people have audiences, the economics around them start to change too. Companies might want to sponsor their projects, give them early access to new tools, or pay to have talented builders create with their products. The thing attracting the audience isn’t content about the work. It’s the work itself.

That’s a very different kind of social network… one built around what people can actually do.

All my best,

This week’s updates focused on making project generation more reliable, responsive, and easier to recover when something goes wrong.

💯 More reliable project generation

• Project generation now runs on GPT

  We’ve moved the project generator’s core workflows from Claude onto GPT models. This gives us more capacity to support learners generating projects at the same time and creates a stronger foundation for improving project quality.
  

• Better handling when generation gets busy

  The generator now responds more gracefully when demand exceeds available model capacity. Instead of failing immediately, it retries model requests with smarter delays.
  

• Failed generations are easier to recover

  We fixed cases where a project could fail near the end of generation and the Try again button could not resume it.

  

• Long project plans no longer freeze the browser

  Generating a detailed project outline previously risked slowing down or freezing the page for several minutes. Outline updates are now processed more efficiently as they stream in.

🏎️ Smoother generator conversations

• No more repeated questions

  We fixed several cases where the generator could ask the same question twice or repeatedly return to questions the learner had already answered.

• Broken conversations can repair themselves

  Malformed or unusually long chat histories could previously break a generator conversation. The generator can now clean up recoverable chat issues before continuing.

• Clearer feedback when the chat runs into a problem

  Instead of silently failing or leaving the loading spinner running forever, our generator now shows an error and gives the learner a way to continue.

🐛 Bug Fixes

• Fixed long task and quiz content overflowing outside its card

  

• Fixed project content containing `{{` breaking the chat.

• Fixed long Learnlist titles overflowing their available space

  

• Fixed some generated chat messages appearing as gibberish.

⭐ New projects

• Give Hermes Agent Powers with MCP

  Connect your Hermes agent to GitHub and your local filesystem through MCP servers, then watch it pull live data from both to summarize a real repository in a single conversation.
  

• Build and Brew projects:

  • Ship Cloud Projects with a Senior Architect

  • Build AI and Python Projects

❤️ Community

• Released an #inspirational-stories wall highlighting a mix of career switchers, uni students and lifelong learners in the community

  

• Released #pick-a-project to make it even faster for members to find their next project within the community.

  

We’ve also set up a Mac Mini in the office that is now taking on tasks, from carousel generation to bug fixes, day and night!

Please continue to share your feedback and feature requests in our community: https://discord.gg/v9BfqUARq

Happy learning!

For decades, proving your skills has always happened after you do the work.

You study, then you take an exam. You build something, then you write the Medium article. You learn a new tool, then you design the LinkedIn carousel.

This sucks. Doing the work and documenting the work has required two different mindsets; you might be a great backend engineer and then have to spend hours trying to make your personal website “look good”. Historically, the need to document your work shows up right when motivation is lowest, just as you’re ready to move on. The unfortunate outcome is that many talented people have very little public evidence of what they’re capable of - not because they lack skills, but because they lack documentation.

NextWork’s true transformation comes from moving learning and proof from separate, to one.

We’re building NextWork to be the Granola of skills. Starting a NextWork session will mean you can build wherever, whatever, and NextWork captures your process and automatically generates a verified portfolio entry as you build. By the time you’re finished, the proof already exists. There’s nothing extra for you to write, upload, or organize.

Before Granola, taking notes was a separate job, or you were stuck with tools that only worked on specific meeting platforms. Now Granola just sits in the background wherever the conversation happens, and the meeting creates its own record automatically. Same thing with NextWork and proof of skills.

Take it one step further - every NextWork portfolio entry can be turned into a tutorial. Another learner can recreate it, extend it, and generate proof of their own. The learner becomes the teacher.

We’re already seeing early versions of this behaviour. Learners aren’t just finding NextWork through our social media channels, they’re finding projects shared by other learners, rebuilding them step by step, and then publishing their own versions. The proof doesn’t just demonstrate learning, it becomes the mechanism by which learning spreads.

The long-term opportunity at NextWork isn’t just to build the world’s best place to learn. It’s to become the infrastructure for human skills.

All my best,

This week’s updates focused on making project creation faster, smoother, and easier to start.

Faster project creation

• Instant Match is now live!

  

  If your prompt for generating a project closely matches an existing project, we’ll now return the ready-made project instead of generating something new from scratch. This should dramatically speed up how quickly learners are landing on project they can start instantly.

• Question cards in the generator

  

  While generating a project, you’ll now see selectable question cards instead of only free-text prompts, making the generation experience clearer and easier to follow.

Better project-building experience

• Carousel blocks are now supported

  Projects can now include richer visual sections with multiple images and diagram-style content.

• Attachment button moved into the prompt bar
  A small UX improvement that makes attaching files more intuitive.

  

New projects

• Added new ChatGPT-focused learning content:

  

  1. Personalize your ChatGPT

  2. ChatGPT Prompting Hack

     
     

• Released new community and Build & Brew-related projects:

  

  1. Apply for Build & Brew

  2. Ship Like a Startup Team in Austin

  3. All Things DevOps, Security + AI

  4. Portfolio-Ready AI Projects Meetup

  5. Everything Cloud, Data and AI

Please continue to share your feedback and feature requests in our community: https://discord.gg/v9BfqUARq

Happy learning!

For decades, the credibility of an individual’s skills largely came from the brands they were associated with: Harvard, Oxford, MIT, Stanford, Google, McKinsey, etc. These institutions spent decades building trust, and individuals borrowed that trust as a signal of what they could probably do.

AI is starting to break that system. Skills are changing faster than institutions can define, teach, and credential them, so the distance between having a credential and actually being capable is getting wider. The credibility that once belonged to institutions is starting to move toward the individual.

This is why I think the idea that AI will transform education by giving everyone a personal tutor misses half the problem. Learning has to get dramatically better, but the real blocker is the bridge between the individual and the rest of the world: a way to prove what you can actually do. AI solves education by forcing the world to create a new proof-of-skills solution.

If the old system was credentials and institutional brand, the new one will increasingly be personal brand backed by proof of work. Instead of saying where you studied or which certificate you earned, you can show what you’ve actually built, how you built it, and how your abilities have developed over time.

If credentials were a way of borrowing credibility from an institution, portfolios give people the ability to build that credibility themselves.

This has already been normal for a small number of professions. Designers have portfolios. Engineers have GitHub. Writers have bodies of published work. In these fields, seeing someone’s work has always told you far more than seeing a certificate saying they can do it.

AI makes this relevant to almost everyone. It isn’t a skill confined to one profession, industry, country, or age group. As AI changes how millions of people work, proving your skills through your work goes from something a few professions care about to something almost everyone will need.

The problem is that the infrastructure for this doesn’t really exist yet. People use LinkedIn posts, YouTube videos, Instagram reels and X threads to show what they’re learning and building, but these platforms were designed for publishing content, not proving capability. The person who is best at showing what they can do shouldn’t have to also be the best at making content about it.

That’s the gap that NextWork is building towards - a global proof-of-skills network that’s:

1. Instant

2. Beautiful to use

3. Focused on real learning instead of fake signals (exams, certs, badges)

4. Supported by a built-in community experience

All my best,

Here’s how that idea became NextWork, and the people, decisions and moments that shaped it along the way.

Amber was born in the UK to a British father and a Kiwi mother, and moved to New Zealand at two. As a kid she signed up for everything she could get her hands on: piano, guitar, darts, tap dancing, taekwondo, art groups, sports. There was endless curiosity about everything, and the belief that you can achieve anything you set your mind to.

The argument (2015)

Amber was finishing high school in Palmerston North, New Zealand, where she had spent most of her time painting and reading. She walked into the lounge one afternoon and found her younger brother, Martin, hunched over a laptop working on a computer science assignment. She asked him to teach her…

He said no, more than once. He told her it was not for her, that she would not enjoy it, that she should not bother.

She took this personally enough that she went into school the next morning and convinced the digital technologies teacher to let her join the senior class despite having missed four years of prerequisites. It helped that the class had six students and could not really afford to turn anyone away.

Amber says she was not particularly good at it. What she found instead was that code worked the way painting did. You could make something out of nothing, on your own, with a strong enough internal sense of what was worth making. But unlike a painting, which hangs on a wall waiting for someone to walk past, code could travel. The same act of creation could reach people you had never met, in places you had never been.

Singapore (2020)

Amber studied software engineering at Victoria University in Wellington and spent a year of it at the National University of Singapore (the highest-ranked university in Asia).

She arrived expecting to travel, explore Asia, and enjoy a year abroad. Instead, she found herself surrounded by students working at a level she had never seen before. Classes where the bell curve sat above 95%. People who had been studying maths and computer science since they were kids. An 85% could feel like failing.

It was confronting, but it changed her sense of what was possible.

Amber came back to New Zealand for her final year feeling like everything was moving in slow motion. Singapore had taught her that her capacity was much higher than she had assumed. The gap between where she was and where she wanted to be was often just how long she was willing to work at something until that level became normal, which was a lesson that would matter almost immediately.

Salesforce Training (2021)

In Amber’s final year of university, she was put into a group project to build a real product for a real client. The client wanted it built in Salesforce. Nobody in the group knew what Salesforce was, and trying to figure it out introduced Amber to an entire industry her software engineering degree had never mentioned.

To learn more about Salesforce, Amber went to a local meetup in Wellington expecting just three or four people… finding 60 or 70 instead. People started offering her jobs after seeing her keep showing up.

The Salesforce world was bigger than anything she had been told about in her classes, and most of the people in it were earning a living from a category of work that nobody had thought to mention to a software engineering student.

She submitted her own honours thesis proposal that year, the only student in her cohort to do so, and used it to build an app for the Salesforce platform and try to sell it. The app was called Salesforce Training, lived on a $20 domain, and did not work particularly well. It was, however, the launching pad for everything after.

A month or two before she was due to start a graduate role at Salesforce, a message came through the site from a New Zealand company asking her rate to train their staff. It was Mediaworks, who ran most of the country’s radio stations. She named a number, took the contract, and discovered there was real demand for someone who could teach a difficult platform to people who had not grown up writing code.

That contract led to a second, the second to a third. By the end of the year she was running a full B2B training business under her own name.

The B2C side grew out of it. People started coming to her individually, not because an employer had sent them but because they wanted to be there. They wanted a promotion, a change, something new for themselves. It was the first time she saw up close what learning looks like when the motivation is internal rather than imposed. It became the thing she would eventually build a company around.

The stakes for these students were rarely small. Amber remembers people working through lunch breaks, logging on before work and after it, and showing up on weekends despite everything else going on in their lives. One student was a former truck driver who had been paralysed and was rebuilding a career from scratch. Another was pregnant and pushing through the course anyway. Another had just lost a parent and was using the moment to change the direction of their life. Amber took the responsibility of getting these people real results personally, and it shaped what she built next more than the certifications ever did.

To go all-in on B2C, Amber put together a free Salesforce workshop that she posted about on LinkedIn. She expected maybe two people to sign up. The post went viral instead, comments piling up with people asking to be enrolled, and around 300 people showed up to the call.

Feeling the momentum and excitement in the call, Amber ended up training a cohort of 16 people over six months, split roughly a third theory, a third hands-on work, and a third toward getting certified.

The certification prep was where most of the course’s time went, but Amber started noticing that the cert didn’t actually help learners get their next job or promotion. The certificate wasn’t the credential anyone on the other side of the table actually cared about.

So what Amber tried instead was capstone-style case studies, real design and architecture problems built around companies she already had relationships with through her training work. She swapped exam prep time for these projects and told learners to talk about the project in interviews instead of the certificate. The results changed almost immediately: learners started getting the job, the promotion, or the move into the team they actually wanted. So Amber leaned all the way into projects and started helping people learn how to talk about what they had built. She would read through drafts of how learners described their projects, cut anything that read poorly, and rewrite it into something that actually landed. That one-on-one editing eventually became a repeatable template, and was the first version of what would later become NextWork’s project documentation.

By two and a half years out of university she was running a profitable services business turning over $700k a year, teaching cloud and Salesforce to hundreds of people, with no co-founders, no investors, and no team. She also had a sister company, Cloud Station, hiring her top students and placing them on contract, which had grown into a small consultancy of its own.

That was the version of the business she could have continued for another decade and made her financially comfortable. It would have been a real company by most reasonable definitions, making the pivot to NextWork all the more powerful.

The pivot (2023)

The pivot from a Salesforce training business to NextWork was triggered by three forces all happening at once…

First, her students kept asking her to teach things beyond Salesforce, AWS and Azure and the wider cloud landscape, and the brand could not stretch to cover it.

Second, Salesforce’s legal team sent an email pointing out that the name Salesforce Training was a copyright infringement… (fair enough)

Third, and most decisive, Amber wrote down a list of everything she wanted to do in her lifetime and realised the list was long enough that she would never get through it on her current trajectory. The services business was making good money, but every year inside it was a year not spent on the larger problem that had been forming in her head through all the manual teaching.

That problem was that online learning did not work. It was boring, lonely, and the completion rates were terrible. It was disconnected from any trust signal an employer would respect. There was nowhere in the world someone could go to learn a tangible vocational skill, prove they had learned it, and walk into an industry that took the proof seriously.

She could see exactly what needed building. She just could not build it while running calls every weeknight from 5:30 to 9:30pm and teaching through weekends.

This is what led her to venture capital - VC investment is not a reward, a credential, or a signal of having made it. It is a direct exchange of money for time. Bootstrapping the platform would have taken two more years of running the services business in parallel. Taking investment bought those two years back.

The beta launch (November 2023)

Amber raised a pre-seed in July 2023, put her own money from the services business into an initial team, and we shipped the first version of the product the same week the funding landed.

Jon joined as founding designer in May, before the pre-seed round even closed. The first team, Amber, Jon, Nigel, Kevin and Nat, gathered on 16 August 2023 for a kickoff and targeted 20 November, three months out, for a closed beta with 70 university students learning AWS, Azure or Salesforce.

That first version of NextWork was essentially a blog. Amber and Nat wrote all the content by hand and Amber was on camera for all of it. The 70 students loved it.

It was also structured like a course:

• Four weeks of theory

• Three weeks of hands-on projects

• Two weeks of certification exam prep

• Two weeks of interview prep.

And time after time, learners raved about the projects the most. That was the part that taught them the most, kept them engaged, and made the theory finally click.

Going live (April 2024)

The beta taught us that what learners needed was different from what we had built. There were also hard lessons about leadership, hiring, and managing expectations.

Amber let the first engineer go and spent four months looking for replacements, sending around eight hundred cold LinkedIn messages and interviewing hundreds of people. Caldas and Krishna came out of that process, and within three months of starting, they shipped a second version that was meaningfully different from the first.

But during those four months without a platform, we started publishing free hands-on AWS project guides on a Webflow page, purely to get attention for a course we planned to sell later.

(Little did we know, the projects became the core of what NextWork is today.)

The community grew alongside it. Rather than asking people to sign up and wait for an email, we opened a forum on Circle, and it became the support layer holding learners through technical problems and the question of which project to start with. It also became the space for learners to show their love for NextWork, which became clear to us that the hands-on projects were really solving a problem that learners were having around finding quality content. People started running events, making videos and even baking cakes about NextWork - completely on their own.

We also started posting on social in April, mostly because it was the only acquisition channel that costs nothing to grow organically. Our first 40 videos didn’t get much attention, but one about a hands-on project hit 10k views on TikTok overnight and eventually crossed 100k.

Maximus joined in June 2024 and took over social shortly. The goal was ten pieces of content a week; at its peak Maximus hit thirty. By the end of the year around 80% of traffic was organic social, with zero paid acquisition.

2025

If 2024 proved the model, 2025 was about scale and momentum:

• Q1 brought a full rebrand, colour themes for documentation, a made-for-you recommendation feature, quizzes, and a rebuilt search and explore experience. Maya joined in March to run community, finishing a postdoc in the early mornings and weekends while balancing NextWork.

• In Q2, Amber and Nat went to SF and raised a $4.5M seed on the back of 90,000 signups across 180 countries, a customer acquisition cost of $0.52 (against an industry average of $116), and a whole lot of learner love.

  

• In Q3, Krishna rebuilt the entire front end in React in a matter of weeks and we shipped a new landing page. Nat also went back on releasing projects alongside an incredible intern, Kahu.

• In Q4, Pano joined, and we released ASK, which lets learners chat with any part of a project, and live documentation, which lets them watch their write-up build as they work rather than only at the end.

2026

If 2025 was the year NextWork was loved, 2026 is the year we become undeniable.

The team relocated from Wellington to Austin in March, saying goodbye to loved ones and taking on a new life in America.

Within a couple months of moving, we also started serving 200,000+ learners, and released new features that let learners generate their own projects. William joined in June as the first engineer hired in person in Austin, after a three-month search.

Then, Jamal and Rita joined in July!

As an ex-NBA coach, Jamal joined to take the helm on social content - and started owning all shorts and carousels within two weeks of starting.

Rita joined NextWork from Corgi’s founding GTM team in Austin, and is now taking NextWork to universities all over the US via a new community program (Build & Brew) she’s leading from scratch.

It was also a huge win to have Loheth and Brandon come on board in August - two incredibly talented and driven engineers passionate about learning and solving tough problems. For the first time in NextWork history, we onboarded two people on the same day!

Where this goes

We want to build a world where learning a serious, life-changing skill doesn’t depend on where you were born, whether you can move across the country, or whether you can afford to spend years waiting for the right institution to let you in.

Someone who wakes up one morning wanting to become a dentist should be able to start that journey that same day. The same should eventually be true for renewable energy, advanced manufacturing, healthcare, nuclear fusion, and every other field where the world needs more capable people.

We know we can’t start there. Before NextWork can help train someone to work on a power grid or care for a patient, we have to earn that trust by proving, at scale, that we can teach difficult skills exceptionally well.

That’s why every new domain matters. Each one is another step toward building an institution people can trust with increasingly important parts of their lives.

We’re building NextWork with that timescale in mind. We want it to still exist in a hundred years, helping generations of people learn what they’re capable of and giving them a way into work that might otherwise have felt out of reach.

We’ve written the first few chapters.

What will we get to write together?

Projects

• Saturday 17th: Build Multi-Region Apps on AWS

• Sunday 18th: Build an AI Security Guard for AWS

• Monday 19th: Build Instant Failover with CloudFront Origin Groups

• Tuesday 20th: Multi-Cloud Disaster Recovery with Pulumi

• Wednesday 21st: Build Your First Azure Function

• Thursday 22nd: Build AI Chat Moderator with Gemini

• Friday 23rd: Orchestrating Parallel Event Streams with Azure

• 21 projects in 21 days nearing completion; last Azure project shipped (noted as very challenging, queued for clarity updates)

• Standardised project intro patterns defined (30‑second summary, track selection, “what you’ll build”), with detailed guidelines for titles, bullets, emojis, and structure

• Ongoing quality pass across projects (intros, steps, policies) to bring them to a consistent standard

Features

• Added back in download markdown options (by popular demand!)

• Automatic OS-aware instructions: Mac/Windows tabs now auto-select based on the user’s operating system, with a fallback to manual tabs when detection fails

Marketing

• x7 Youtube videos

• x7 Shorts

• x7 Emails

• x7 Landing Pages

• x7 Project Step Carousels

• Internal “nano banana”/Napkin‑based image generator prototype for on‑brand diagrams and assets (Markdown + brand‑guidelines aware)

• Mobile signup modal removed to reduce friction; follow‑up analysis planned on impact to sign‑ups and re‑engagement

Community

• 5 community calls and 2 pre‑built labs run over the weekend

• New build labs delivered, including Krishna’s Azure lab (called out as “next level”)

• Project Spotlight launched with community members (e.g. Okasha)

• Organic learner‑led labs emerging (e.g. Azure project club, DevOps + AI series)

• New community members joining and immediately engaging via live sessions (e.g. Germany, San Diego)

• Early city/community group experiments underway (towards “50 groups in 50 states”)

• New record: 30 hours of live community time this week

Quality of Life fixes

• Multiple quality-of-life fixes and code cleanup: PDF download bugs fixed, ~2k lines of unused code removed, and React modal migration underway (expected to remove another 2–3k lines)

• Editor no longer needs a full page refresh to show saved content (removing the need to spam “Reset changes”)

• YouTube comment sync live: pulls new comments ~every 3 hours (≈12 runs/day) into a single internal discord channel with direct “Watch video” links. Goal is to increase the replies we can give to our wider community, create a great experience/interaction with NextWork.

Projects

• Sunday: Build an AI Security Scanner with Gemini (Part 3)

• Monday: Deploy Your RAG API to Kubernetes (Part 4)

• Tuesday: Automate RAG Testing with GitHub Actions

• Wednesday: Ship a Payment Landing Page with v0 + Vercel

• Thursday: Secure Payments with Antigravity + Stripe

• Friday: Payment Analytics with PostHog

• Saturday: Build Multi-Region Apps on AWS

• New YouTube Playlists:

  • 21 projects in 21 days

  • DevOps + AI series

  • All AI Projects

Project Automation

• Automated image framing in app

• Image placeholders

• Validation checkers to avoid errors in mismatched number of validations between tracks.

Features

• Preview feature improvements to the social sharing part of MyWork

  

• Last Updated timestamp for projects

  

  
  

Growth

• x7 Youtube videos

• x7 Shorts

• x7 Emails

• x7 Landing Pages

• x7 Project Step Carousels

• x3 Project Architecture Diagram Carousels (new!)

• x3 Project Definition Carousels (new!)

Community

• x7 BuildLab events

• x1 Humans of NextWork event with Krishna

• x5 Connect with Community events

• Improved social sharing assets for community events (new!)

Projects

Goal: 21 projects in 21 days.

• Day 1/21. Released 9th Jan: Build a RAG API with FastAPI
  Build your very first AI API. Use Python and FastAPI to answer questions with your own files - no experience needed.

Project Automation

Goal: enable learners to generate NextWork projects on the fly.

• Internal Claude skill for project documentation

• Internal Claude skill for reviewing content repo

• Internal Claude skill for creating validations

Features

• Pro plan removed, using NextWork is entirely free.

Growth

Goal: x14 pieces of quality content per day.

• x2 Youtube videos; Project Live, Project Demo

• x1 Short on Facebook, Instagram, LinkedIn, Tiktok

• x1 Email

• x1 Landing Page

• x1 Project Step Carousels on Facebook, Instagram, LinkedIn, Tiktok

Community

• x3 BuildLab event in Discord, live on YouTube and Twitch

• x1 Humans of NextWork event with Maximus

• x5 Connect with Community events

• Events now also on Luma.

Learner Love

Navigation on the Explore menu was improved by adding categories and filters for Roadmaps, Specialties, and Tools.

• Added keyboard shortcuts for opening the Explore Menu and the Portfolio.

• Started the React migration. This took majority of our engineering capacity this week as we’re working on moving a significant amount of our app to React from Web Components. We’re doing this so we can make use of front-end libraries for animation and UX experience, and design faster using LLMs and AI native building tools. We expect this to be a three week sprint, followed by work on the ASK and Live Documentation features!

• Highlight menu changes direction if it looks like it’s going off screen

  

• Reset project also resets the documentation generated, allowing students to generate multiple documentations per project.

• The prompt behind ‘Made for You’ is updated to give better, more accurate results for students.

• New roadmaps out for:

  • AI/ML Roadmap

  

  • DevSecOps Roadmap

    

  • Solutions Architect Roadmap

    

• Roadmap landing page for Cloud Engineer live!

• Add link to easily unsubscribe or change membership plan

🔹 Improvements

• Filter indicator to show what status filter is applied on My Projects

  

• Saved user variables are now available for inline code blocks

  

• Community homepage update

  

• Pro badge now shows in Explore and in My Projects

• Updates to support AI native browsers such as Dia

• Add animations to transitions for expanding/hiding steps

This update delivers comprehensive event tracking across the platform, enhanced bookmark functionality, improved Circle Login experience, and support for educational document management. We've also implemented automated email notifications for new users and project completions.

✨ New

• Note edit adds a blank line on typing

🔹 Improvements

• Improved Circle Login Page - Improve the Circle community login page

• Bookmark on callouts

• Adjust crop on search results details line

• Update Cloud Security with IAM project from student feedback

• Comprehensive event tracking implementation, capturing key user interactions including: • Project navigation (restart, next, add projects) • Documentation interactions (viewing, saving as README/Markdown) • Pro feature engagement (unlocking content, clicking on pro-related elements) • UI interactions (expanding/collapsing sections, menu interactions) • User journey events (portfolio access, project completion)

• Updated Amazon Lex 3 project

🚀 New Features

• Send an email on first project completion

• Send an email 2 days after signup

• BUILD Bookmark - Bookmark feature

The past few cycles were brief and disrupted by the Easter and ANZAC holidays. However, we successfully delivered comprehensive documentation improvements, including enhanced pattern documentation and formatting. Additionally, we made refinements to the user interface, such as font upgrades and the introduction of the new AI Hallucinator Indicator for reporting.

The consolidated implementation of the Pro badge and the improved project components streamline the user experience. Furthermore, several critical fixes have been addressed, including issues with profile refresh and document sharing functionality.

✨ New

• AI Hallucinator Indicator

• Consolidate where the Pro badge is being used

🔹 Improvements

• LinkedIn sharing URL

• Tidy Up for Documentation - Documentation Formatting still getting complaints

• Update whole project, not just one track

• Font Upgrade across all components across design

• Floating checklist formatting

• Consolidate all components into a UI kit, including Master and Child components for project contents and documentation for all patterns (both primary and nested)

• PRO badge when you're PRO

• Add where documents are being shared to, via the sharing modal

• [UPDATE] Amazon Lex 2 project

🧪 Testing

• n/a

🐞 Fixes

• Documentation Formatting still getting complaints

• Profile refresh does not refreshes nickname and profile pictures

• PDF Generation failure status

• Internal links not working

• Markdown document images broken on GitHub

Cycle 13 delivers the the anticipated Restart Project feature, giving users the ability to reset and begin projects fresh. Yew! Additionally, we've introduced project revamps and updates, UI improvements include multiline checkbox positioning and profile refresh updates, while notable fixes address LinkedIn sharing functionality and document rendering issues.

✨ New

• Restart Project; Pro Feature

• Getting Started with AWS Projects [RELEASED]

• Create the single completion email template

• Current selected project made private
  

🔹 Improvements

• Pro cancel issues resolved

• Live dash reload

• Check if nickname is empty on profile refresh

• Add job to recreate user nicknames with problematic content

• All tracks saving

• Accept numbers on variable names

• Single completion emails sent to students

• Update reset to trigger intro refetch

• Add what the user asked, to the analytics event

• Multiline checkbox positioning
  

🧪 Testing

• Visualize data with QuickSight

• Network Series Intro
  

🐞 Fixes

• Current selected project made private

• Email single project completion to students

• Revert "Fix LinkedIn sharing functionality due to LinkedIn changing their api."

• Fix LinkedIn sharing functionality due to LinkedIn changing their api.

• Remove hidden responses from doc

• Validation hidden attribute

• Fix scroll_to_sentry errors

• Table of Content: Fix markers not showing as focused when they are in view & delete the old one

• Adding Variables to Code Blocks
  Often in a NextWork project you’ll need to add a new variable directly into a block of code - from simple things like your initials, to AWS account IDs, to regions, and more. Now you can do this directly in NextWork…and we’ll remember your variables too! Even across multiple projects! 🥳
  

  

  
  
  Previously, the process to do this was to copy the code, paste it somewhere else, update your unique variable, then copy and paste the final version into your terminal. No more.

New Projects (x1)

• New project and YouTube live on Build a Security Monitoring System on AWS

New Videos (x20+)

• Over 20 new videos released across YouTube, Instagram, Tiktok, Facebook, and X.

• Made For You tile

  Finding the right projects for your goals can be tricky! Now with Made For You, you can enter your goal and we’ll generate a list of NextWork projects that best meet your needs.

  

  
  

• Quiz Me feature

  Test your knowledge at the end of a project with the Quiz Me feature. Questions and answers are generated using AI based on every individual project.

  

  
  

New Projects (x8)

• New project on Join the 7 Day DevOps Challenge!

• New project and YouTube live on 7-Day DevOps Challenge (DAY #1) | Set Up a Web App in the Cloud

• New project and YouTube live out on 7-Day DevOps Challenge (DAY #2) | Connect a GitHub Repo with AWS

• New project and YouTube live on 7-Day DevOps Challenge (DAY #3) | Store Dependencies in CodeArtifact

• New project and YouTube live out on 7-Day DevOps Challenge (DAY #4) | Package an App with CodeBuild

• New project and YouTube live on 7-Day DevOps Challenge (DAY #5) | Deploy an App with CodeDeploy

• New project and YouTube live out on 7-Day DevOps Challenge (DAY #6) | Automate with CloudFormation

• New project and YouTube live on 7-Day DevOps Challenge (DAY #7) | Build a CI/CD Pipeline with CodePipeline

Polish

• Fixed documentation formatting

  We fixed a minor bug that made some of the documentation generated format in strange ways. All better!
  

• Up to 1000 characters.

  Our Pro users can now enjoy up to 1000 characters while crafting their task answers.
  This will allow them to create even more detailed documentation of their projects.

  

  
  

New Projects (2x)

• New project and YouTube live on Build a Web App for your RAG Chatbot

• New project and YouTube live out on Secure Secrets with Secrets Manager.

Polish

• Update our checkout experience.

  Fixed issue to avoid delays on Pro subscription.

• Fixed LinkedIn share documentation.

  We fixed a minor bug preventing users from uploading their documentation while sharing on LinkedIn via our share button.
  

• In Progress Status + Filter
  View which projects are not-started, in-progress, and complete. Filter by project status on your My Projects page.

  

  

  
  

• Show Parts of a Series
  Showing parts of a series within the Explore tab to make it easier to navigate and understand which projects are stand-alone and which are part of a larger series.

New Projects (x2)

• New project and YouTube live out on P2 - Create an API for Your RAG Chatbot

• Community events in Vancouver and virtually with @ShaeInTheCloud

Polish

• Update Grey Swatch
  A little left over colour work from our re-brand at the start of the year - updating all our hover and grey swatch colours.

• x3 Colour Themes for Documentation
  Choose from one of three themes for your documentation; light, dark, or fun. It’s the start of more personalised ways to showcase your skills.

  

• Add custom URL to the title page
  Add a URL to your LinkedIn, GitHub, or personal website directly on the title page of your documentation.

New Projects (x2)

• New project and YouTube live out on P1 - Set Up a RAG Chatbot in Bedrock

• New project and YouTube live out on P2 - Chat With Your Bot in the Terminal

• Community events in Lagos and Atlanta.

Polish

• Adding tiles to the Explore menu
  Now it’s easier (and more fun!) to find the different series of projects.

  

• Remember your tab selection; either `Explore` or `My Projects`
  NextWork will remember which tab you were on last and open you to that tab directly next time you search for a project.

  

Fixes

• Notes being generated in some callouts as plain text
  We noticed that some notes were being generated as plain text and displayed inline with the project content. This is fixed! Let us know if you see anything suspicious 🤨