January 19, 2026

Adam’s 2025 Year in Review

Adam Cogan AI, General, SSW Projects 1 Comment

For us Aussies, December ended in a way that changed the tone of the whole year. The Bondi Beach mass shooting targeted a Jewish celebration and killed 15 people. This one was close to home (literally), as I live a couple of beaches down from Bondi and work with some Jewish team members at SSW. All Aussies hate this stuff coming to our world, and it made Christmas more sombre than usual.

It reinforced something I’ve felt all year… when the world feels chaotic, what matters most is community. I’m really proud of how Aussies came together in the aftermath to support each other.

While the end of 2025 was a horrible shock, I’d love to focus on some of the positives from the year.

For me, 2025 was the year AI went from “interesting” to indispensable. I saw it become part of every software project we did and every role at SSW. I saw people automate the boring stuff, create crazy smart workflows, get insights into their recorded meetings, and of course, write too much of their code…

Here’s a look back at the highlights: the family moments, the travel, and what we’ve built at SSW – I feel that we’ve genuinely moved the needle.

My 2025 highlights

My 2025 highlight… I couldn’t decide on one… so I’ve given you my top 6.

🧑‍💻 All of the European NDCs! We travelled with TinaCMS and introduced NDC Norway & NDC Copenhagen to our cool AI bug reporting tool “YakShaver.”

🐂 Seeing the growth of YakShaver, which went from a fun internal SSW tool, to a client-facing productivity multiplier. What started as “make bug reporting less painful” turned into an AI tool that everyone who uses it loves it.

🦙 SSW Rules being migrated to TinaCMS – I’m so proud of how good the UX is now for users – smooth as butter! Keep an eye on my blog next month for the story behind the migration. 👀

☀️ I loved going to Spain for my friend’s 50th celebration. Eddie Geller was once a Microsoft consulting competitor in Sydney and now runs a cool startup in NYC called Getskor.com – Imagine a massive house full of funny and smart characters, the conversations were awesome. 🥳

🧑‍💻 The Microsoft MVP Summit in Seattle was awesome. I love seeing all my RD and MVP friends. I learned a lot.

🐼 Visiting SSW China for 3 weeks, that office is always so welcoming, and the SSW team took turns taking us out to do a lot of fun things! Thanks to Jimmy, Clara & Alvin for making it happen!

**Figure: The SSW French team and I enjoyed hanging out at NDC Oslo** **and working the booth** **– we had some great successes and learnings**

**Figure: Each day I did a prize draw, developers love fluffy llamas!** **SSW’s booth was busy**.

Highlights from SSWers

The entire company did a retro of 2025, and I wanted to share some of the highlights. There was a lot of excitement about successful client projects, new tech, buying houses and travelling.

Here’s some of my favourites…

🤖 Winning many new AI projects + Coding with AI & Knox + YakShaver MCP

🗣️ Got personally called out by a client in a company-wide email for going above and beyond to deliver a $2 million project… Great team – lucky to have an amazing Product Owner and great colleagues such as Toby

🎓 Working on a huge project with Dan and Gordon – Absolute weapons and great mentors

🥖 Strasbourg – spent 5 months in the French office, thanks to the SSW Cultural Exchange

🏠 Helped to fully pay off the mortgage for my parents’ home!

🎵 This year, I produced and played a lot of music, and I enjoyed having a hobby that gets me out of the office chair.

**Figure: Richard Campbell did an awesome talk at our Sydney Brainstorming – it was definitely a highlight of 2025!**

**Figure: Another shot of me enjoying Richard’s talk** **at SSW Sydney**

SSW Projects

2025 was dominated by AI projects, workshops and conversations. Yes, we still had our traditional .NET projects, but what an interesting year for AI projects and developers who are using it in productive ways. The goal for most is to become a 10x developer.

Two that stood out for me were projects that broke the mould and have the potential to make a huge impact in the development space.

Using AI to reduce the cost of legacy code

The first was seeing Solution Architect & Microsoft MVP, Gordon Beeming, using AI to revolutionise the way we deal with legacy code bases.

One of the reasons companies avoid moving away from often end-of-life legacy code bases is the cost involved in rebuilding it. Gordon and his team were able to use AI to look at the codebase and execute 2 massive modernisation projects in record time. For one client, they migrated 8 major Angular version upgrades in less than a day. For another, they compressed a multi-month migration of a .NET 4.7 system to .NET 10 in 2 weeks, while simultaneously modernising legacy views and controllers into React and Minimal APIs in 5 weeks. I recall that not very long ago, a transformation like this would typically be estimated at over 6 months.

Using AI for safety – ‘near miss reporting’ in heavy industry

In heavy industries such as mining, ports, and logistics, radio communications are an important safety input. Teams record vehicle-to-vehicle radio calls and review those recordings to check whether safety protocols were followed. This review work, looking for safety breaches, can take a long time and vary widely from person to person. When people switch off, and key parts of calls are missed, it can mean near-miss incidents (and hopefully nothing worse).

This year SSW worked with a mining client on a feasibility project that added AI into this existing workflow. The prototype takes exported radio audio files, produces text transcripts, and applies automated analysis to flag potential protocol gaps. Reviewers can correct the results and export structured reports. SSW also provided dashboards so teams can see where communications fall short.

The system works with recorded files, not live radio feeds, and it has not been deployed as a formal safety control. It was built to assist people in their existing process, rather than replace them.

One outcome we observed is that reviewer corrections can be used to improve future model performance. Tracking these corrections over time, can help the analysis become more accurate for different accents, vocabulary sets, and radio formats. This makes the approach easier to roll out across sites.

AI in this case, helped reduce manual effort and improve consistency in the review process. I am interested in seeing how similar approaches could support safety reporting in other industries.