Articles about LLMs and agents from the ConsultorIA blog: cloud, applied AI and security in production.
The cost per token looks obviously cheaper until you account for utilisation. GPUs bill by the hour whether or not anyone is asking questions, and that single fact decides most of these decisions.
Any agent that reads untrusted content and can act will eventually take instructions from that content. The defences that work are architectural, not textual.
A classifier on every request costs time and money on every request. What to check on input, what to check on output, and how to decide whether the guardrail is worth what it takes.
A hundred real cases, per-task metrics instead of one score, and a judge calibrated against human labels. That is enough to catch regressions and to decide whether a model change is safe.
It is not a certification of your model. It is the scaffolding that makes AI governance repeatable, and if you already run ISO 27001, most of the machinery is built.
An approval step is the cheapest safety control and the easiest one to place badly. Put it too early and you have automated nothing; too late and you are rubber-stamping. Four placements and when each is right.
Fine-tuning teaches behaviour and format. Retrieval supplies knowledge. Prompting shapes the immediate task. Most teams reach for the expensive one to fix a problem it cannot fix.
Most teams either assume the AI Act does not touch them or assume it bans what they are building. How to establish your role and risk tier, and what the obligations cost an engineering team.
Traditional monitoring tells you the workflow succeeded. With a model in the middle, succeeding and being right are different things. What to capture, and the three dashboards worth building.
The Model Context Protocol turns an internal API into something an agent can call. The protocol is the easy part; the design decisions are which operations to expose, at what granularity, and under whose identity.
Teams argue about which model to use and then ship an assistant that answers confidently from the wrong document. The decisions that actually decide quality are about chunks, permissions, evaluation and cost.
Most of the friction people blame on the model is a missing paragraph in the repository. What belongs in a context file, what does not, and why the ones that grow past a page stop working.
The bottleneck moved from writing to reviewing, and the old review habits do not scale to it. Four gates that keep quality without turning one senior engineer into a queue.
One keeps you in the loop at every keystroke, the other takes a task away and brings back a diff. Most teams that argue about which to standardise on end up using both, for reasons that are easy to state.
A coding agent with shell access and repo context is a different tool from an autocomplete. The gains are real and they land unevenly; the cost is that review becomes the bottleneck. What we changed after six months.
An open-source assistant that lives on your own server, talks to you through the messaging app you already use and runs tools on your behalf. Genuinely useful, and the most over-permissioned thing most people will ever install.
An AI node in a workflow is a non-deterministic step in a deterministic pipeline. That is fine for classification and drafting, and dangerous for routing and writes. Here is where we draw the line.
A prompt is production configuration that changes behaviour silently. Treat it like code, and build the smallest evaluation that can tell you whether a change made things better.
Oracle's managed model service covers the plumbing for a standard assistant, including a vector-capable database you probably already have. Here is the architecture and the decisions it does not make for you.