Automation

AI Agents for Business

An AI agent is a workflow that decides its own next step. It reads a request, looks things up in your systems, chooses an action and takes it. That is genuinely useful for a narrow, well-bounded job, and genuinely dangerous when the boundaries are vague, so most of the work in building one is deciding what it may not do.

  • Hard limits on every action an agent can take
  • Shadow mode before it touches anything live
  • Full log of every decision

Who this is for

Businesses that have already automated the fixed-rule work and are left with tasks a person handles by checking two or three systems and using judgement: qualifying an enquiry against eligibility rules, chasing a document, reconciling an order against a payment, preparing a quote from a specification. This page sits under marketing and business automation services, and it is the most advanced form of the work described on AI automation services.

If you have not yet automated the deterministic part of the process, an agent is the wrong purchase. Agents are more expensive to build, harder to test and slower to trust than a rules-based workflow, and they should only take over the part that genuinely needs judgement. Start with workflow automation and add an agent to the step that rules cannot express.

Agent, chatbot or workflow

WorkflowChatbotAgent
Decides its own stepsNo, the path is fixedOnly within a conversationYes, within the tools it is given
Typical jobForm to CRM to WhatsAppAnswering visitorsQualify, look up, prepare, act
Testing effortLowMediumHigh, and never finished
Failure modeStops and alertsSays something wrongDoes something wrong, repeatedly, quickly
Right choice whenThe rules are knownPeople need answersThe judgement is real and the actions are reversible

What I build agents for

  • Enquiry qualification. The agent reads a message in Nepali or English, checks the CRM for an existing record, checks a sheet or system for eligibility, drafts the reply, proposes a next action and books a slot if the person asks for one.
  • Document chasing. For processes that need paperwork, the agent tracks what is outstanding per case, sends the right reminder, reads what comes back, checks it is the right document and updates the record.
  • Order and payment reconciliation. Comparing what the store says, what the bank statement says and what the sheet says, flagging the differences a person should look at rather than fixing them silently.
  • Research and preparation. Assembling a briefing before a sales call from the CRM history, the website, past emails and public sources, so the person walks in prepared.
  • Internal operations. Assigning work by workload and skill, escalating what has stalled and preparing the weekly review, which is the sort of thing built into an internal tool or dashboard.

What an agent is never allowed to do

  • Move money. No payments, refunds, transfers or credit notes. It can prepare one for a person to approve.
  • Send to a customer unreviewed, until the measured error rate on a real test set justifies it, and never for complaints, medical, legal or financial content.
  • Delete anything. Records are archived or flagged, never removed.
  • Act outside its tool list. An agent can only do what the workflow explicitly exposes to it. There is no general access to your systems, no shell, no arbitrary browsing.
  • Run without a budget and a stop. Every agent has a step limit, a token limit, a daily spend ceiling and a pause switch anyone can use.

How a build runs

  1. Write the job description. Literally: what a new employee would be told to do, what they may decide alone, and what they must ask about. If that cannot be written in one page, the task is not ready for an agent.
  2. List the tools and the limits. Every read and every write the agent may perform, with the boundary on each: which records, which fields, which value ranges.
  3. Build the test set. Fifty to a hundred real past cases with the correct outcome, including the awkward ones. The agent is scored against these before anyone sees it, and rescored after every change.
  4. Shadow mode for two to four weeks. The agent proposes; a person still does the work; we compare every decision. Nothing goes live until the comparison is boring.
  5. Limited live running on the lowest-risk slice of cases, with every action logged and reversible, and an alert when the agent takes an unusual path.
  6. Widen slowly, guided by measured accuracy, not by enthusiasm. Reviews are quarterly, because model updates change behaviour and the test set has to be re-run.

What you receive

  • The agent running in your own automation account, on your own model API keys.
  • The written job description, tool list and limits, in language your team can audit.
  • The scored test set and the accuracy record, so changes are measured rather than assumed.
  • A full decision log: what the agent saw, what it chose, what it did, and how to reverse it.
  • Spend ceilings, alerts and a pause switch, plus documentation and a handover session.

Pricing pointer

Agent setup is quoted on request and sits at the upper end of automation builds because testing, shadow running and guardrails are most of the effort, not the model call. Maintenance and quarterly accuracy reviews are quoted on request, and model usage is billed to your own account. If what you actually need is a team assistant that answers from your documents rather than one that takes actions, that is cheaper and lives on custom GPT and AI assistant development.

Next step

Bring one task, the systems it touches and twenty real past cases. I will tell you within four business hours whether it needs an agent, a plain workflow or neither, and what the guardrails would have to be. I turn this work down more often than I take it, which is the point.

Frequently asked questions

What is the difference between an AI agent and a chatbot?

A chatbot converses and answers. An agent takes actions in your systems: it reads records, decides a next step and performs it. That makes an agent more useful and more consequential, so it gets a defined tool list, hard limits, a scored test set and a shadow-running period before it touches anything real.

What happens when the agent gets it wrong?

Every action is logged and reversible by design, and anything irreversible is outside the agent's tool list. Unusual paths raise an alert while it runs, and the case goes back into the test set so the same mistake is measured rather than argued about. Early on, a person reviews everything, which is how errors get caught before they cost anything.

Can an agent talk to customers directly?

Only after the measured error rate justifies it, and never for complaints, medical, legal or financial subjects. The usual arrangement is that the agent drafts and a person sends, then the review rate is reduced for the categories where accuracy is proven. Direct customer contact is a decision you make on evidence, not a default setting.

How much does the model usage cost?

More than a simple workflow, because an agent may make several model calls per case as it checks and decides. The estimate is produced from your case volume during scoping, with a step limit and a daily spend ceiling configured so a loop cannot produce a surprise bill. Usage is billed to your own account, not through me.

How long before it is trustworthy?

Plan on two to four weeks of shadow running after the build, then a limited live period on low-risk cases. The honest answer is that trust is earned by the accuracy record, not by the calendar, and some tasks never reach the point where the review step should be removed. That is a valid outcome, not a failed project.

Do we need an agent at all?

Often not. If the decisions can be written as rules, a plain workflow is cheaper, faster, testable and cannot hallucinate. Agents earn their cost when the input is unstructured, the judgement is real and the same judgement is made many times a week. I say which case you are in before quoting.

Ready to talk about your project?

A free 30-minute call, a straight answer about what would move the numbers, and a written proposal within 48 hours if we are a fit.