AI – Everything Digital Tech and More

AI and your money: this is your heads-up.

April 18, 2026

I sent the below to my newsletter subscribers today 18th April 2026. If you stumble on this, I believe you will find it useful.

For context, my newsletter is focused on Finance for Immigrants

My current role is in Transformative AI. Meaning we are designing and developing solutions that will replace the processes humans are currently carrying out.

This week in particular, we have been in immersive sessions on various AI topics with key leaders, technical and non-technical.

Let me share some things they shared that I think will help improve your professional relevance in the next couple of ~~years~~ months.

AI pioneer Andrew Ng (Google him) made a point that should stop every professional in their tracks: he expects every single employee—whether they are in marketing, finance, or HR—to know how to use AI to write code and automate their workflows. He clarified to say that he didn’t expect them to actually write code like Developers. But you must be able to use AI coding tools.eg Claude Code
The head of my team made it clear to her direct reports if you aren’t actively using AI coding and productivity tools, you aren’t just behind—you are becoming unemployable.
The New Standard for Your CV; If you are an immigrant navigating the UK or global finance markets, “Microsoft Office” is no longer the gold standard. Your CV needs to scream AI-Native. Employers are no longer looking for people who can do the work; they want people who can orchestrate the tools that do the work.

What to add to your CV right now:

Prompt Engineering & Workflow Automation: Evidence of how you’ve used LLMs to cut a 5-hour task down to 15 minutes.
AI-Assisted Analysis: Mention specific tools (like ChatGPT, Claude Code, Gemini) that you use to increase productivity.

We are seeing the birth of entirely new career paths. If you are looking for your next move, look for these titles:

AI Orchestrator: These aren’t just “managers.” They are professionals who know how to stitch different AI agents together to complete a complex business process from start to finish.
Agentic Outcome Managers: People who don’t manage humans, but manage “Agentic Workflows.” You define the business goal, and you manage the AI agents tasked with hitting that outcome.
AI Evaluation (Eval) Managers: This is a massive gap in the market. As businesses deploy more AI, they need people to “watch the watchers.” You will be responsible for testing, auditing, and ensuring the outputs of AI agents are accurate, ethical, and financially sound.
Red Teaming Specialists — actively stress-testing AI systems to find risks, biases, and failure points. Google McKinsey AI hack and Code Wall for example.
AI Guardrails SMEs — defining the boundaries, controls, and policies that keep AI systems safe and compliant.

These roles are not all for developers. A non-technical person can do them. Please spend time upskilling and gain experience volunteering for projects.

The Bottom Line: In 2026, you don’t need to be a Computer Science graduate to “code.” You need to be an AI-literate professional who isn’t afraid to use a prompt to build a solution. The barrier to entry has lowered, but the standard for productivity has skyrocketed.

Take action my fellow immigrant.

Data & AI Series; Without the Jargon

November 16, 2025

(An attempt at educating non-techies and Business decision makers on AI-related topics)

I’ve been immersed in data and AI discussions over the last couple of years, through hands-on projects and continuous self-learning. It’s clear that every business problem these days seems to require an AI solution. It’s the hottest topic everywhere, and because an “AI project” sounds innovative, many leaders are pushing for “AI in all we do.”

However, the more time I spend in this space, the clearer it becomes that AI isn’t the solution every business problem (or process) actually needs.

When You DON’T Need AI

If your process is repetitive, predictable, rule-based, structured, or happens the same way every time—it’s an automation problem. Examples of Automation Problems:

Pulling weekly reports from a fixed set of sources.
Notifying teams when a ticket changes status (e.g., from ‘New’ to ‘Assigned’).
Assigning tasks based on fixed departmental rules.
Moving data between two systems in a set format.
Approving expenses that are consistently under a certain monetary threshold.

No AI is required for these tasks. This just needs simple automation and process clarity. I know, “Automation cleanup” sounds boring, but here’s the reality: AI on top of broken processes just speeds up the chaos.

So, When Do I Need AI (Especially Agentic AI)?

AI (and Agentic AI) is useful ONLY when the work involves genuine uncertainty, context, or judgment:

Interpretation: Reading, classifying, or understanding nuanced meaning.
Judgment: Deciding between options based on context, not fixed rules.
Handling Ambiguity: Dealing with inputs that have no clear rules or structure.
Multi-step Reasoning: Planning and executing multiple sequential steps.
Dynamic Adaptation: Adjusting behavior when situations change live.

Agentic AI is just a term for AI that can plan, act, adapt, and complete tasks end-to-end—it doesn’t just answer questions; it drives to a conclusion.

Here are the questions to ask before proposing an AI solution:

Is the task repetitive and predictable? → If yes, automation, not AI.

Are there clear rules you can write down?→ If yes, automation.

Does the task require interpretation or judgment?→ If yes, you’re leaning toward AI.

Does the process depend on context or changing conditions?→ Agentic AI might be helpful.

Does the task involve multiple steps that change depending on outcomes?→ Consider agentic AI workflows.

I am the first to admit AI is powerful. But using AI for every problem is like hiring a lawyer to assemble IKEA furniture—it’s expensive overkill. As a business leader, you need to match the right tool to the right problem. Often, the smartest AI strategy… is no AI at all.

Happy to answer any questions you might have. If I don’t know the answer, I’ll happily consult my network of experienced colleagues and friends to get you the right information.

xxxxxxx

Anu Sanya