Every AI employee we build is custom. After enough hires, patterns emerge. Here are the six roles we ship most often, plus a seventh path: something built entirely from scratch for a job no template covers.
An AI employee who knows your product as well as your best support agent, available the second a customer hits your help page.
Trained on your past tickets, FAQs, product docs, and (where helpful) your live order or account database. They answer in your voice, escalate the right tickets to a human, and tag everything so your team can see exactly what they’re doing.
Best for: SaaS products, e-commerce stores, service businesses with repeat questions about scheduling, returns, account access, etc.
Typical lift: 60-80% of incoming tickets handled end-to-end.
An AI teammate who drops into your contact form, live chat, or website and runs the discovery conversation your sales team would, if they had time.
Asks the qualifying questions, scores fit against your ICP, captures contact details, and books a meeting on a calendar only when the lead clears your bar. Hands you a clean summary in Slack or your CRM.
Best for: agencies, consultancies, B2B SaaS, anyone whose inbound forms get a lot of noise mixed with the real opportunities.
Typical lift: 3-5x more qualified meetings booked, half the time wasted on bad-fit calls.
A private Slack or Teams teammate who answers internal questions before someone has to interrupt your senior team to find the answer.
Trained on your SOPs, runbooks, contracts, onboarding docs, internal wiki, and recorded all-hands. New hires get up to speed faster, your seniors stop being human search engines, and institutional knowledge stops walking out the door when people leave.
Best for: teams over 15 people, anyone with a sprawling Notion or Confluence, businesses with technical procedures that take time to learn.
Typical lift: roughly 4-6 hours/week reclaimed per senior team member.
An AI drafter who produces first drafts of the documents you write the same way every time, in your voice.
Quotes, proposals, scope-of-work docs, follow-up emails, statements of work, summary memos. Trained on your past documents and house style guide so the output reads like you wrote it. You review, edit, send.
Best for: agencies, law firms, consultancies, sales teams writing repetitive proposals.
Typical lift: 80% time reduction on first drafts. Two-hour proposal becomes a 20-minute review.
An AI ops employee with permission to actually do things in the tools you use, not just answer questions about them.
Hooked into Stripe, HubSpot, QuickBooks, Notion, Google Workspace, your custom backend. You ask in plain English: refund a customer, send an invoice, schedule a follow-up, pull a report, update a record. They do it, confirm what they did, and log everything for audit.
Best for: ops-heavy businesses, founders doing the admin themselves, teams drowning in copy-paste between tools.
Typical lift: a part-time ops role’s worth of repeated tasks gone.
A phone teammate who picks up when your staff can’t, books appointments, takes orders, or screens calls before routing.
Sounds natural, knows your business, transfers to a human when they should. Captures every conversation as text in your CRM. Works with your existing phone number.
Best for: service businesses (clinics, salons, dispatch, hospitality), anyone losing bookings to voicemail after hours.
Typical lift: after-hours bookings captured at 30-50% of daytime rates instead of zero.
When the job doesn’t match a template, we start from a blank page.
Plenty of the work we ship doesn’t fit the six roles above. A research analyst that reads industry filings every morning and posts the highlights. A scheduling coordinator that juggles a fleet of contractors. A fraud reviewer that flags suspicious orders before they ship. A sales SDR that re-engages stale leads on a 90-day cycle. We’ve built all of those.
If you can describe the job in a paragraph, we can almost certainly build the employee. Same process, same timeline, same fixed-price model. The discovery call is where we figure out whether it’s actually a fit.
Best for: a role that’s specific to your business, your industry, or the way you operate.
Typical engagement: $5k–$25k depending on complexity, integrations, and the depth of training material we’re working with.
Engagements typically run $3k–$25k for a single AI employee, depending on the complexity of the role, integrations needed, and how much custom training material we’re working with. We give you a fixed quote after the discovery call, not before.
Two weeks before your new hire is on the job in some form. Four to six weeks total from kickoff to fully deployed and tuned. Voice teammates and deep CRM integrations sit at the longer end.
You do. The training, the playbook, the code, the deployment. We hand it all over with documentation. If we vanish tomorrow, your AI employee keeps showing up to work.
Optional retainer for monitoring, ongoing training, and new responsibilities. Most clients keep us on for the first three months, then take it in-house. A few keep us on indefinitely as their AI team.
Off-the-shelf chatbots are generic. They follow a script, can’t reason, and give up the moment a question goes off the path. We build employees that actually know your business: trained on your real materials, integrated with your real tools, accountable for outcomes. The difference is night and day.
Your training data stays yours. Used only to teach your AI employee its job, never used to train anyone else’s model. We sign whatever NDAs and DPAs your legal team needs. See our security page for the detail.