# AI agents + automation

Source: https://www.nevdev.io/services/ai-integrations/

A team of AI agents trained on your business, working alongside your staff and managed by them.

## How we approach it

Think of each agent as a new member of the team with one job. It signs into the tools you already use, learns how your business runs, and keeps working while your staff get on with theirs. Your team gives it work the way they would message a colleague, and it comes back to them when a decision needs a person.

- Each agent holds a role: following up leads, clearing the inbox, updating the CRM, preparing reports, chasing invoices
- Trained on your business: your prices, your policies, your tone of voice and the way your team already works
- Agents work together, passing tasks between each other and keeping the context of every conversation
- Nothing important happens without a person: payments, prices, bookings and complaints come back to your team for approval
- Customer service chatbots for Instagram, WhatsApp and your website, answering enquiries day and night
- The last week of every build is spent training your team to manage the agents: what each one does, how to correct it, when to step in

We will tell you when a workflow does not need a model. A rules engine that always behaves the same way is usually cheaper, faster and easier to trust than a model that behaves the same way most of the time.

## What sits inside it

- AI agent teams
- Agents trained on your business
- Lead + inbox agents
- Instagram, WhatsApp + website chatbots
- Workflow automation
- Team training

## Where it has been used

- [Vision Photography](https://www.nevdev.io/work/vision-photography/): A small photography studio answering the same questions all year, and trying to win new work at the same time, with the same few people.
- [Lumeneer](https://www.nevdev.io/work/lumeneer/): The work is finished and the job is not: eight hundred frames still have to be culled, corrected and put somewhere a client can actually use.