BJSS

Thanks to DIGIT.FYI for sharing this article by Valcon on how to measure ROI on AI projects. Whilst Koko (our Valcon Group Head of AI) won't be in attendance at the #AIBusinessSummit2026 , there will be plenty of #AI #expertise from Valcon and others at the #event . Valcon are excited to be the #Headline #Sponsor at this event. I am even more excited because I get to be back in #Glasgow for a whole week with this being the highlight of the week. That and a trip on Wednesday afternoon to meet with Partick Thistle Football Club . I won't say much more than that just, except to wish Partick Thistle Football Club all the best against St Mirren on Thursday, in the 1st leg of the promotion playoffs. Partick Thistle Football Club are my local team, and it would be great to see them back in the Scotland premiership. Historic club, great fans, real community ethos and hopeful ambition.

Last week I had the opportunity to do something that I always wanted to do, but never could.... I was able to offer a local student a one week work experience placement at MCR Agentic . This is something that I've wanted to do for a long time but was always pushed back on by higher ups at companies that I worked for. Turns out, when you're the head of a Microsoft Partner, you can bring someone on for work experience whenever you want! It was a great week, where the student learned lots about AI tools available, building out 2 websites using Lovable. He built one website to help students study for their GCSE's, and another about ex-Man United players and what they were up to now. We were clear that these are great for Proof of Concepts and not necessarily something you could actually publish, but he really learned a lot about how everything fits together, and where you can use AI. It was a great experience for me too, particularly to be able to give back and support Manchester based students learn a bit more what the world of work is like, and some of the tools that we use day to day. I'm really looking forward to hosting another student again in the future

How good is your cyber security vs OpenAI? A great example of cyber maturity being measured in ability to respond, not in resilience or posture. Just like in sport, at crunch time you fall back to your training - and having response plans that have been documented, tested, and shared is key. Third party risk management is only one half of this story, incident response is what saved the day today. - Credentials rotation - Revoke user sessions - Infected service isolation https://lnkd.in/ejxk6bxG

Let's talk about AI and low performers. AI is giving them cover and it's going to bite companies in the ass. Here's the problem: Low performers generate everything with AI now. Decks, emails, analysis, code. It looks intelligent. Sounds smart. Passes the first review. But when pressed, they can't: -Think -Ideate -Explain their own work -Function as subject matter experts Because they aren't experts. They're mimics. AI did the work. They just took credit. And the worst part is most managers can't tell the difference. The output looks good so they assume competence. Until the project goes sideways. Until the client asks hard questions. Until you need someone who actually knows what they're doing. Then you realize: You've been paying someone to be a middleman between you and ChatGPT. If you can't think without AI, you can't think with it either.

Found myself at the #AWSSummit in London yesterday with the Capgemini Invent team. A brilliant event with so much going on! A few talks really peaked my interest: - How Climate X are able to provide incredible data to their clients, allowing them to accurately quantify climate related risks and take appropriate action. - Skyscanner ’s evolving infrastructure architecture, from a server in their founders bedroom, to now scaling to ~200m monthly users and billions of requests in AWS. - How Luminance were able to work with Amazon Web Services (AWS) to leverage their own legal-specific LLM and the challenges and successes they found when deploying into the cloud. #AWS #Capgemini

I recently built SayIt, a voice-first Android task management prototype designed to reduce friction in everyday task capture. Most task apps still assume users want to type, tap through forms, and manually manage lists. I wanted to explore a simpler question: What if a daily task app worked at the speed of a spoken thought? So I built an Android prototype that: opens directly on today’s task list supports voice commands to add, complete, query, and delete tasks includes a monthly calendar view for browsing tasks by date uses simple visual states to show whether a day is completed, overdue, or still active A few product decisions I focused on: Today-first navigation instead of burying daily work behind a calendar Voice as the primary input, not just an extra feature Explicit listening / processing states so the system feels responsive Low-friction gestures, like double-tapping blank space to start listening Calendar status colors to make progress visible without opening each day Tech stack: Kotlin Jetpack Compose Room Android TextToSpeech Android system speech recognition One important product insight from building this: the fastest prototype path is not always the best production path. The current version uses Android system speech recognition, which made prototyping fast, but it also exposed a real constraint for users in mainland China, where Google-backed recognition may be unreliable or slow. That made speech infrastructure not just a technical detail, but a core product decision. So the next step is clear: evaluate domestic cloud SDKs vs offline ASR keep the voice-first UX while improving production reliability This project was a useful reminder that good mobile UX is often about removing interaction cost, not adding more features. GitHub: https://lnkd.in/eWQDNhDH If useful, I can also share the case study / slide breakdown behind this project. #Android #Kotlin #JetpackCompose #MobileApp #UXDesign #ProductDesign #VoiceUI #SpeechRecognition #CaseStudy #PortfolioProject

One of the most satisfying parts of what we do at Ingenaium is being able to solve some of our clients hardest problems very quickly. Everyone is curious about AI, but it is not a tooling question - it is about identifying the right use case, showing how it could be solved quickly, and then understanding what this really means to an organisation. Business owners can articulate their problems quickly. We can show how they may be solved quickly via the Ingenaium platform. Then, we move forward together. If you've got potential use cases, we've got the time to discuss and show how they may be solved. The clients we already work with are seeing considerable ROI via sensible application of AI to their business in a way that can be measured - please reach out if you want a chat.

Next stop: Orlando ✈️ After two delays, finally on my way to ConnexFM . Looking forward to connecting, learning, and seeing everyone! :D See you there. 😁🙏⛱️🌞

One thing I’ve come to value in technical leadership is the ability to reduce ambiguity. In many projects, delivery does not become difficult because the engineering is impossible. It becomes difficult because ownership is unclear, priorities compete, and teams are asked to move quickly without enough alignment. Over time, I’ve found that a good Tech Lead helps by creating clarity around a few essential things: - what problem the team is solving - which trade-offs matter most - where quality is non-negotiable - what is good enough for the current stage That does not mean being involved in every detail. If anything, it often means the opposite: creating the conditions for the team to move confidently without unnecessary noise. In my experience, strong technical leadership often shows up quietly through clearer decisions, better sequencing, and more reliable delivery. That remains one of the most important parts of the role for me. #TechLead #SoftwareDelivery #EngineeringLeadership