The home of AI Integrators
Operandis matches growing companies with vetted AI Integrators: finance, operations, sales and service professionals who automate the processes they already master. Proof of work included.
New role. New standard. First cohort open for Portugal and Spain, remote across Europe.
Every mid-sized company hits the same wall.
Orders go up, so you hire in operations. Invoices go up, so you hire in finance. Tickets go up, so you hire in customer service. Each hire takes three months to find, three months to train, and buys you one more year before the next one.
Meanwhile your team spends its days moving information from one place to another: copying data between systems, chasing suppliers, answering the same twenty questions, building the same report every Monday.
You know AI could do most of that. You have seen it on your own laptop. But an AI engineer does not understand your invoices, a consultancy leaves after the slide deck, and nobody on your team has the time to learn.
The missing piece is not a tool. It is a person.
Illustrative shape, not data: every new layer of work buys another hire.
“We did not need another pair of hands. We needed someone who could make the hands we had go three times further.”
An illustrative statement of the problem, not a client quote.
A professional from your sector who learned to build with AI, and now rebuilds the processes they already know.
An AI Integrator is not a developer and not a consultant. She has spent years in finance, or operations, or sales, or service. She knows which steps are pointless, which handoffs break, which reports nobody reads. Then she learned the tools: automation platforms, AI models, internal tool builders, agents.
Now she sits inside your company and turns the process she understands into a system that runs itself, one process at a time.
The title always carries the function: Finance AI Integrator, Operations AI Integrator, Sales AI Integrator, Customer Service AI Integrator.
Days 1 to 15
She shadows the team, lists every recurring process, and measures the time each one takes. You get a ranked list of what to automate first, with hours attached.
Days 16 to 45
Quick wins with the biggest hours. Live, in production, with the team using them.
Days 46 to 90
Documentation, monitoring, training. The team knows how each automation works and what to do when it fails.
Illustrative scenarios of typical builds. Not client results.
Finance
An agent that reads supplier invoices from the inbox, extracts the data, matches them to purchase orders and bank movements, and queues exceptions for a human.
Operations
A workflow that takes incoming orders from email, web and marketplace, normalises them, checks stock, and creates the fulfilment task with the right priority.
Customer service
A first-response assistant that drafts answers from the company's own policies and order data, checked by a human before sending, and fully automatic for the 15 most common questions.
Sales
A pipeline assistant that enriches every new lead, drafts the first follow-up in the rep's voice, and flags deals that have gone quiet.
Logistics
A shipment monitor that reads carrier updates, predicts delays, and notifies the customer and the account manager before anyone asks.
HR
An onboarding flow that generates contracts from templates, collects documents, schedules the first week and answers new-hire questions from the handbook.
One person. Five processes. The difference between a team of twelve that feels like twelve, and a team of twelve that feels like thirty.
Why they move so fast
The people we place understand how a business really works: the processes, the exceptions, the reason behind every step. Until recently that knowledge stopped at a ticket for IT, a request to a developer who did not quite get it, or a spreadsheet that only went so far.
That limit is gone. With today's AI tools, someone who understands the process can observe it, redesign it and build the improvement themselves, in days. No translation layer, no waiting list. That is why they move so fast, and why the company moves with them.
Before AI
With AI
Four steps. No retainers, no slide decks.
What you get
Three minutes. Tell us what you do, how big you are, and which processes hurt. We read every application ourselves.
Forty-five minutes with Jesus or Alex. We explain what an AI Integrator is, map which of your processes an Integrator would attack first, and tell you honestly whether this role fits you now. No charge, no obligation.
Within two to three weeks, two to four vetted Integrators from your sector. Each profile comes with the automations they have actually built, the before-and-after numbers they measured, and our vetting notes.
Hire them on your payroll, or through an employer-of-record service if they live in another country. We stay close for the first 90 days and step in if the match is not working.
If you have spent years in your function and taught yourself to build with AI, you are an AI Integrator. Companies are looking for you under the wrong title.
You are the one in your company who keeps saying "we could automate this". You built your first tools on your own time, you use AI every day, at work and at home, and lately you feel the people around you are not moving at your pace. Is this you?
Apply as an IntegratorWho should apply
What you get
Domain depth
You know your function from the inside: its edge cases, its bad days, its month-end.
AI fluency
You build with current tools, know what a model is good and bad at, and design for the failures.
Proof of work
Real automations, in production, with measured results. We weight this above everything else.
Plain speech
You can explain a process and an automation to a company owner without a single acronym.
Every Integrator on Operandis has passed the same standard. Here it is, in full. We do not certify people on courses or self-assessments: we verify three things, in this order, and a candidate must pass all three.
A structured interview with a founder or a vetted Integrator from the same function. We go deep on how the work really happens: what breaks, what the team hides, what month-end feels like. We check that the candidate has lived the function, not studied it.
A review of two real automations the candidate has built: architecture, exception handling, data flow, how credentials are kept, and how a non-technical colleague would use it. Live walkthrough, screen shared, questions answered.
The candidate presents one integration with a measured outcome, and we verify it: how it was measured, over what period, and what else changed at the same time. We prefer honest small numbers over impressive vague ones.
A four-hour take-home built on a realistic process from the candidate's sector, using their own tools. We are not grading perfection. We watch how they scope, what they automate first, and how they explain it to the owner.
Pass rate target: we expect to accept fewer than one in five applicants. The standard is the product.
Accepted: fewer than 1 in 5
Example profiles. Operandis is new. These four cards are illustrative composites of the profile we vet for, not real candidates. Real profiles will replace them as the first cohort is accepted.
Measured (illustrative)
Month-end close from 10 days to 3.
Measured (illustrative)
Order re-keying eliminated; delay calls down by half.
Measured (illustrative)
First response from 6 hours to under 1.
Measured (illustrative)
Rep admin time from 40% to under 15% of the week.
Operandis does not have a wall of client logos yet. We are starting now. What we can promise is how we work: we vet every candidate ourselves, we take every orientation call ourselves, and we publish our standard in full.
When the first Integrators are placed, their results will appear here, with their permission and with real numbers.
Operandis is a project by Jesus and Alex.
Where we are today
Simple and success-based. When an Integrator joins your payroll, you pay a one-time placement fee on their start date. If you prefer, the same fee can be spread as a monthly fee while they work with you, which suits companies hiring someone in another country through an employer-of-record service. No retainers, no fees for the shortlist. We walk you through both options on the orientation call.
An AI engineer knows models and code, but not your invoices, your suppliers or your customers. They need someone to tell them what to build, and that someone is usually too busy. A consultancy maps your processes, hands you a deck, and leaves before anything runs. An AI Integrator comes from your function, decides what to build, builds it, and stays to make it work. One person, inside the company, who owns the outcome.
Both work. Most Integrators work remotely with regular on-site time; we recommend at least one week on-site in the first month so they can shadow the team. If you want someone fully on-site, we tell you upfront if that narrows the shortlist. The first cohort is based in Europe.
From orientation call to shortlist, two to three weeks. From shortlist to start, it depends on the candidate's notice period: usually between two and eight weeks.
Every placement includes a replacement period, agreed in writing before you hire. If the match is not working, we tell each other early and we find a replacement at no extra placement fee. Our vetting is designed to make this rare, and our 90-day check-ins are designed to catch it early.
Integrators work inside your systems, under your access controls and your policies, like any employee. We sign a confidentiality agreement with you before the orientation call goes into detail, and every Integrator signs one with you before day one. During the Proof of Build we check that candidates handle credentials and personal data properly: it is a pass or fail criterion, not a nice-to-have.
Not "uses ChatGPT". We mean someone who can take a real process, decide which steps a model should handle and which a person must keep, build the workflow with current tools, design for the cases where the model is wrong, measure the result, and explain the whole thing to a non-technical colleague. That is what the Proof of Build and the practical exercise test.
Whether you are a company that needs to move faster, or a professional who already builds, the first cohort is open now.
Every application is read by Jesus or Alex. You will hear back within five working days.