Swiss-based technology advisory firm, specialising in Software Due Diligence and Tech Advisory. https://teapawt.net/t.co/LET5ZaDFS2
Switzerland
Joined March 2018
- Tweets491
- Following280
- Followers96
- Likes112
Most organisations already know the baseline for AI agents: least-privilege access, human sign-off on risky actions, logging, and a way to shut them down. The NCSC’s recent post covers this well.
How much autonomy do AI agents have in your organisation?
ncsc.gov.uk/blogs/managing-t…
Two productive companies do not automatically make one.
After an acquisition, teams still need to deliver while connecting systems and adapting to shared processes. The integration itself takes capacity that the growth plan needs to account for.
How does Java support AI in enterprise applications?
Fabio Vellosogave a technical talk on integrating AI models into Java applications and managing their behaviour in production.
avalia.io/post/java-enterpri…
A decade of Avalia Systems. 🎂
What we’ve built comes down to our people: their experience and the work they put in for our clients every day. Thank you to our clients and partners for the trust you place in us. Here’s to the next one!
Happy birthday, Avalia!
DOT Digital Group cut a 224-hour task to 48. Not with a better tool. With a better workflow.
Which process would you redesign from scratch?
224 hours → 48 hours. Same task.
Each square is one hour spent on [the task]. Together with DOT Digital Group, we rebuilt the process around it and gave back 176 of them.
More on this monday!
#AvaliaTechForum #ProcessImprovement
AI pilots are easy to start. Scaling them is harder.
At Avalia Tech Forum 2026 in São Paulo, Ana Paula Gutierrez from DOT Digital Group shared what changes when AI moves into real operations.
We summarized the main points here:
avalia.io/post/scaling-ai-in…
What is your operational efficiency actually resting on?
Southwest found its technology limit by hitting it. The alternative is to go looking on purpose.
avalia.io/post/technology-du…
By the time a software delivery problem shows up as a missed deadline or a budget issue, the warning signs have often been visible for weeks. avalia.io/post/delivery-qual…
1. Waiting: where work sits still: review queues, approvals, blocked releases, dependencies.
2. Rework: what keeps coming back: reopened tickets, hotfixes, rollbacks, repeated fixes.
3. Unplanned work: what starts displacing the plan: incidents, firefighting, production fixes.
Software engineering has never produced more data about itself.
Yet the question that matters most stays hard to answer: is the business actually getting more value from technology?
avalia.io/post/engineering-m…
Engineering metrics can show that delivery is getting faster.
But outside the engineering team, numbers like deployment frequency or lead time often raise another question:
What does this actually change for the business?
Recent research, including Microsoft’s EngThrive framework, reinforces something we see as increasingly important: the same engineering data needs to answer very different questions depending on who is looking at it.
Why technology reporting fails at the board level, how operational metrics obscure business impact, and what boards need to make better investment and risk decisions. 👇
avalia.io/post/why-technolog…
DX Hub makes your portfolio something leaders can actually explore. This 3D view is one way to see it. You spot where complexity concentrates and where action is needed, then take it to the board with evidence.
Most "productivity" gains are just future problems you haven't paid for yet. Is durable productivity a process problem or a culture problem?
youtu.be/b39qApl1lAM
How the right visibility turns engineering reality into the language your board actually acts on.
Worth a read if you've ever left a leadership meeting knowing your teams delivered and wishing you could show it.
avalia.io/post/cto-prove-bus…
