Skip to content
AI & Machine Learning4 min read

AI agents for business: what they can do today and how to deploy them safely

A grounded look at AI agents: the workflows they handle well, where they still fail, and the permissions, review steps and monitoring that make them safe to run.

Engineering & Design

Published
Isometric nodes connected in a loop around a central block on a dark grid

Key takeaways

  • An agent is an LLM that plans steps and calls tools, so its risk comes from what those tools can do.
  • Good first agents work on narrow, repeatable workflows with clear success criteria.
  • Give agents the least permissions possible and require approval for irreversible actions.
  • Log every step so you can review, debug and improve.

"Agents" are the most talked-about idea in AI right now, and also the most loosely defined. Some products use the word for a chatbot with a few buttons. Others mean software that can read your inbox, update your CRM and send emails on its own.

This guide cuts through that. It explains what an agent actually is, which jobs it does well today, and the controls you need before you let one touch real systems. If you're earlier in the journey, start with how to add AI to your product without the hype.

What makes something an agent

A chatbot answers. An agent acts. In practice an agent combines:

  1. A language model that interprets the goal and decides the next step.
  2. Tools it can call: search, database queries, APIs, file operations, a browser.
  3. A loop that feeds each tool result back to the model until the task is done or it stops.
  4. Memory or state so it can keep track of progress across steps.

The model is the planner. The tools are where the real-world impact happens, which is why tool design and permissions matter more than the choice of model.

Where agents work well today

The best candidates share three traits: the task is repeatable, the inputs are available to the agent, and the outcome is easy to check.

WorkflowWhat the agent doesHuman role
Support triageReads new tickets, tags them, pulls account context, drafts a replyReviews and sends
Lead researchLooks up a company, summarises fit, fills CRM fieldsApproves before outreach
Back-office reconciliationMatches records across two systems, flags mismatchesResolves exceptions
Report preparationQueries data, builds a summary, highlights changesChecks and shares
Internal knowledge tasksFinds policies or past decisions and cites themUses the answer
Document processingExtracts fields and routes documents, see AI document extractionHandles low-confidence cases

Where they still struggle

  • Long, open-ended tasks. Errors compound across many steps.
  • Ambiguous goals. "Grow our sales" is not a task; "draft follow-ups for leads with no reply in 14 days" is.
  • Untrusted inputs. Emails, web pages and documents can contain instructions that try to hijack the agent (prompt injection).
  • Irreversible actions. Payments, deletions and messages to customers need a human checkpoint.
  • Cost and latency. Every reasoning step is a model call; long loops get slow and expensive.

Designing an agent: a checklist

Scope the job

Write the task as a one-sentence instruction with a measurable outcome. Define what "done" means and what the agent should do when it can't finish.

Design narrow tools

Prefer get_open_tickets(customer_id) over a generic "run any SQL" tool. Narrow tools are easier to secure, test and explain.

Apply least privilege

  • Use dedicated service accounts with only the permissions the task needs.
  • Default to read-only. Add write access one action at a time.
  • Scope data access to what the user who triggered the agent could see.

Put humans at the right checkpoints

Require approval for anything that sends, pays, deletes, publishes or changes permissions. Show the human exactly what will happen before they confirm.

Treat inputs as data, not instructions

Content the agent reads (emails, web pages, uploaded files) should never be able to change its goal or unlock tools. Keep system instructions separate, and filter or flag suspicious content.

Log everything

Record each step, tool call, input and output. Logs make debugging possible and give you an audit trail.

Evaluate before and after launch

Build a test set of real tasks with expected outcomes, score every change against it, and add failures from production. The approach mirrors the evaluation section in our AI product guide.

Build, buy or combine

SituationUsually best
Common workflow in a popular SaaS toolUse the tool's built-in AI features
Workflow spans your own systems and rulesCustom agent with narrow tools
Answers over your documentsRetrieval first; see RAG vs fine-tuning
High-risk domain (finance, health, legal)Human-in-the-loop by design, with expert review

Rolling out

  1. Shadow mode. The agent runs alongside people and its output is compared, not used.
  2. Assisted mode. The agent drafts; people approve every action.
  3. Partial autonomy. Low-risk actions run automatically; the rest still need approval.
  4. Review regularly. Check success rates, costs and incidents, and tighten or relax permissions based on evidence.

Where to start

Pick one workflow your team repeats every day, with clear inputs and an easy-to-check result. Build it in assisted mode and measure time saved.

Our AI development team designs and builds agents with the permissions, evaluation and logging described here. Tell us about the workflow you'd like to hand off.

Frequently asked questions

What is an AI agent?

An AI agent is a system where a language model decides which steps to take and which tools to call, such as search, a database query or an API, to complete a goal. Unlike a chatbot, it acts as well as answers.

What business tasks are AI agents good at?

Multi-step tasks with clear inputs and checkable outputs: triaging support tickets, researching and summarising leads, reconciling records, preparing reports, and drafting responses that a person approves.

Are AI agents safe to give access to company systems?

They can be, with guardrails: least-privilege credentials, read-only access where possible, human approval for actions that send, pay or delete, strict scoping of what data they can see, and full logging.

Do we need a custom agent or an off-the-shelf tool?

If an existing tool covers the workflow and your data, start there. Custom agents make sense when the workflow depends on your internal systems, data or rules that off-the-shelf tools can't reach.

How do you measure whether an agent is working?

Track task success rate on a fixed test set, the share of runs needing human correction, time saved per task, cost per completed task, and any incidents.

  • AI agents
  • LLMs
  • Automation

Written by

Engineering & Design

Notes from the designers and engineers building products at BreaktroughF1: what we learned, what we'd do again and what we'd do differently.

Keep reading

Isometric document sheets feeding into a grid of data cells on a light background
AI & Machine Learning

Automating invoice and document data extraction with AI

How modern OCR and language models pull structured data out of invoices, receipts, forms and IDs — and how to design the validation and review steps that make it reliable.

BreaktroughF1 Team · 3 min read
A neural network diagram of connected nodes on a dark grid
AI & Machine Learning

How to add AI to your product without the hype

A practical framework for choosing AI features that users actually value — and shipping them reliably, safely and on budget.

Noeal Rajeev Thaleeparambil · 6 min read

Newsletter

Get new articles in your inbox.

Practical notes on building software, hardware and products. No spam — unsubscribe anytime.