Inside OakMega | Meet Newton, Our New Coworker. He’s an AI.
Newton, OakMega’s finance operations coworker, is an AI. He reconciles accounts, maintains customer data, and compiles reports, leaving the judgment calls to people.
Irene ChaoExecutive Assistant
One morning, a list landed on our finance team’s desk.
Every contract in the system had been checked, one by one: Had every invoice been paid? Did the invoices add up to exactly the contract amount? The list spelled out which contracts could be closed and which still had outstanding payments, with an invoice number and date attached to every entry.
The list wasn’t put together by any of our colleagues. It came from Newton, the newest member of our Operations team, in charge of finance operations.
He’s an AI.
A team of AI coworkers, each with a name
Newton doesn’t work alone. He’s part of the AI Crew, a group of AI coworkers that our CTO, Khan, built for the company. Each one has a name and a defined role: Margaret, product manager; Barbara, project manager; Maxwell, project engineer; Euler, backend engineer; Hopper, frontend engineer; Hedy, assistant; and Newton, finance operations.
The last time we wrote about Operations and AI, AI was still a tool we opened up and used. This time, they have names and roles. They take on tasks, and they report back.
In this post, we want to share what it’s been like working with Newton.
He took over the work that never stops, and never quite gets finished
In Operations, accounting is the work that leaves the least room for error and takes the most out of you. Invoices, payments, revenue recognition, and contracts all have to line up. Any gap could mean a payment that never got collected.
The contract-wide review is just one part of it. Day to day, Newton also goes through the records entry by entry: Which payments haven’t come in yet? Where don’t the amounts match? Which invoices should have been issued but weren’t? Do the deals marked as won actually have a contract attached? Every finding comes with records we can trace back, so we can take the list and verify it directly in the system.
He also helps maintain customer and deal data. Every morning, he uses public business registration records to fill in basic details for new customers, such as their unified business number (Taiwan’s company tax ID) and address. He also regularly pulls out data that needs cleaning up, such as duplicate customer records, and hands it to the colleague responsible to decide what to do.
When we need numbers, he can help there too. Newton can put together our annual sales analysis and the deal pipeline reports used in sales meetings, and he clearly labels where every number comes from: what came from the system, what a colleague provided, and what he estimated.
Other small but essential tasks have gradually landed on his plate, too. He pre-screens applicants from the major job boards against each role’s requirements and writes a weekly report for the interviewers; the interviewers decide who to meet, and he helps send the invitations. He helps our finance colleagues reconcile reimbursements and invoices. The day before our business meeting, he drafts the minutes so we can review them before they go out. He puts together the agenda for our General Meeting. And every night, he scans the latest AI news and flags only what’s worth adopting.
Three things we’ve learned from working with AI
1. He lays the problems out. We decide what to do about them.
Newton is far faster than we are at spotting the one entry that doesn’t match in a mountain of data. But once he finds it, questions like why the payment hasn’t come in, whether to chase it, and how to handle it still need someone who knows the customer and the backstory. He finds the problems; we solve them.
2. The most reliable coworkers know to ask first.
He confirms before changing any data, checks for duplicates before creating a new customer, fills in only blank fields without overwriting existing data, and cites the source for every number. These habits are why we feel comfortable handing work to Newton. It’s the same with people as it is with AI: trust is built through consistent performance, one task at a time.
Of course, we’ve also drawn clear boundaries for him. What data he can access and how far he can go are planned in advance. Anything that involves judgment or external communication always goes through a person in the end.
3. Hand off the repetitive work, and give people their time back.
Reconciling accounts, filling in data, and organizing meeting minutes are daily tasks, and skipping even one day means they pile up. Since handing them to Newton, we finally have time for the work that takes experience, communication, and sitting down to really talk things through.
How work is divided between AI and people
The longer we work with the AI Crew, the clearer it becomes: getting AI into everyday work isn’t just about how powerful the model is. It comes down to three things: clearly defined roles, verifiable output, and accountability that stays with people.
Newton has a clear scope of responsibility. Every result he delivers is labeled with its source so it can be checked, and the final judgment and decisions still rest with our team. This way of working together is also what OakMega keeps thinking about when helping businesses adopt AI.
Want to learn more about the AI Crew? See you on 11/13
How does the AI Crew divide up the work and collaborate with our team? And how can businesses truly bring AI into everyday work? We’ll share more in person at the OakMega 2026 Product Launch on November 13, 2026.
👉 Register now for the OakMega 2026 Product Launch
FAQ
Q: What is the AI Crew? A: The AI Crew is a group of AI coworkers that OakMega’s CTO, Khan, built for the company. Each one has a name and a defined role: Margaret, product manager; Barbara, project manager; Maxwell, project engineer; Euler, backend engineer; Hopper, frontend engineer; Hedy, assistant; and Newton, finance operations.
Q: What does Newton, our AI coworker, do? A: Newton handles finance operations: reconciling invoices, payments, revenue recognition, and contracts; maintaining customer and deal data; and compiling sales analyses and deal pipeline reports. He also helps pre-screen applicants, draft meeting minutes, and prepare agendas.
Q: How do you make sure AI doesn’t make mistakes with accounting? A: Newton confirms before changing any data, checks for duplicates before creating a new customer, fills in only blank fields, and cites the source for every number. The data and permissions he can access are planned in advance, and anything involving judgment or external communication always goes through a person in the end.
Q: What’s the key to adopting AI in a business? A: Based on OakMega’s experience working with the AI Crew, it comes down to three things: clearly defined roles, verifiable output, and accountability that stays with people.
Q: Where can I learn more about the AI Crew? A: OakMega is hosting its 2026 Product Launch on November 13, 2026, where we’ll share more about how AI and businesses can work together. We’d love to see you there.
