Using Simply in Recruitment

What does Simply actually do? An honest overview

Simply is an AI-native ATS and CRM for recruitment teams. That’s the short version, and the important word in it is the second one, because for a long time this page said something else. Simply used to be a layer you bolted onto the ATS you already had. It isn’t any more. It is the system of record itself.

Concretely: your candidates, jobs, clients, applications and placements live in Simply, in a data model you define yourself. Conversations are recorded and summarised against those records. CVs are parsed into them. Matching runs over them. Client-ready documents are generated out of them. Dashboards count them. There is no second system and no sync in the middle.

That sounds like a feature list. It comes down to one thing: the admin stops being a separate job. Not because a tool does it in another window, but because the place the work happens and the place the record lives are the same place.

The problem Simply solves

As a recruiter you spend most of your day on tasks that have nothing to do with recruiting. After every call you type a summary. You open the record and fill in fields by hand. A CV arrives, you open Word, you start reformatting. You switch between four or five tools. Tab, tab, copy, paste, save.

That isn’t just frustrating. It costs you placements, because while you’re doing admin a competitor is already calling that candidate back. Speed is a competitive advantage in recruitment. Every minute on admin is a minute not spent talking to people.

Research puts recruiters at around 62% of the working day on administrative tasks. On eight hours, that’s close to five.

The interesting question is why so much software has failed to fix it, and the answer is structural. Nearly all of that admin is data entry into the system of record. A clever tool sitting next to the system of record can read a conversation beautifully and still has to hand its output over a connector, into fields it does not own, in formats it has to guess at, with a mapping project attached and a sync to babysit. The work does not disappear. It changes shape and moves.

Simply removes the hop instead of optimising it. The AI and the record are in the same system, so an extracted value lands in the field it belongs in without crossing anything.

How Simply works in practice

Theory is nice. Here’s a typical working day.

Step 1: Record conversations, on any channel

You start with an intake over Google Meet. Then you call a candidate back on her mobile. After lunch there’s a client meeting on Teams. In between, a quick call from the car.

Recording covers all of it. Online meetings are captured natively through Google and Microsoft, which means nothing dials into your call as a guest and no candidate has to ask who that fifth attendee is. Phone calls run through the desktop app, in-person conversations through the mobile app. All of it arrives in the same place: on the candidate record, next to their messages in the unified inbox.

That last part matters more than it sounds. What was said in March is still there in June, in the same context the AI reasons over, rather than in a transcript archive nobody opens twice.

Step 2: Summaries that match the conversation type

Within a couple of minutes you have a summary. Not a generic rundown of everything that was said, which is where most tools stop.

AI summaries take their shape from the conversation. A screening call comes out as a screening summary. An intake comes out as a brief: role requirements, team, culture, terms, selection process, dealbreakers. A debrief comes out as a recommendation with the reservation attached rather than sanded off. Each one carries its action items and the field updates it wants to make.

Formats are configurable per conversation type and per client. Set it up once and every summary for that client looks the same, whoever ran the call. That’s not only speed. It’s consistency across a team where experience varies.

Step 3: Data entry that arrives as a proposal

This is the least glamorous part and the one every recruiter names first when you ask what ruins their day.

A conversation, a message or a document produces proposed changes on the record they belong to. A notice period. A new mobile number. A rate that moved. Each proposal carries a confidence score and the line it came from, and you approve or reject it. Nothing is written to a record without a human approving it, and what you approved, and that it was you, goes into the audit log in the same transaction as the value itself.

It fills real fields rather than a notes blob: dropdowns selected from the options that exist in your model, dates in the right format, numbers that are numbers. Because you defined those fields, there’s no mapping layer trying to work out what your system calls things.

Fifteen calls a day at eight to twelve fields each is 120 to 180 fields. Approving a list is not the same job as typing them.

Step 4: CVs in, client documents out

A candidate sends a CV in whatever format they have. PDF, Word, sometimes a LinkedIn export. CV parsing reads it against your data model: work history and education become linked child records with their own structure, languages come out at CEFR level, and every employer named is resolved against a company database. Every row carries its source and a confidence between 0 and 1, and the rules omit rather than guess. If a CV never states a notice period, the notice period stays empty instead of becoming a plausible number somebody later quotes to a client.

Out the other side, document generation takes a candidate, your template and the vacancy they were matched to, so the opening paragraph argues for this person against that role instead of opening with three sentences about a driven professional. Tone, target language, bullets and anonymisation are choices at generation time. After that you edit in plain language, styling included.

Step 5: Matching, and being able to show your work

AI matching scores one candidate against one job, and one job against your whole database. The score is the weight of the requirements a candidate satisfies divided by the weight of everything the vacancy asked for, so you can recalculate it by hand if a client challenges it. Every score says which lane produced it: a model, or the formula. And it shows the part that doesn’t match, which is the half most tools leave out.

All of it is traceable. Every sentence in a transcript is anchored to the moment in the audio it came from, so you click a line and hear it said. Every AI output says whether it came from a model or a fixed rule. Every write is in an append-only log. When someone asks what a decision was based on, that’s an answer rather than an argument.

Dashboards sit on top, over the objects in your own model. They return aggregates and stop there: counts, rates and trends, with no click-through from a chart to the rows behind it. Where the pipeline sits, what converts, where placements come from. Not a per-person scorecard.

What systems does Simply work with?

Two honest answers, and the first one is the one people are surprised by.

Simply does not sync with another ATS. Not Bullhorn, not Carerix, not anything in that category. There is no write-back and no two-way connector. Simply is the system of record, and two systems that are both the truth drift apart. Moving to Simply is a one-way migration: four weeks, with a data audit in week one so you know before you commit what your export carries across and what it doesn’t.

For the rest of your stack, there are six connectors and an API. The connectors are Google Calendar, Gmail, Outlook, Slack, Zoom and HubSpot. Six, named, all of them built. You click through the OAuth flow once for the workspace and the tokens are refreshed for you from then on. There is no logo wall with “and 300 more” underneath it, because that is a marketing surface rather than a list you can check in a demo.

Anything outside those six goes through the public API, which is generated from the same metadata model the interface runs on. Your schema is your API surface: define an object, add a field, and it’s in the responses without a release from us. Keys are per workspace, outbound webhooks are HMAC-signed so your receiver can prove they came from us, and there’s an MCP server if you want Claude Desktop or Cursor working against your own data.

Who is Simply built for?

Simply isn’t for everyone, and that’s deliberate.

It’s built for professional recruitment teams running volume:

• Staffing agencies doing dozens of conversations a day who need fast, accurate records.
• Secondment firms who need structured candidate profiles for their end clients.
• Search firms who present CVs in a house style and keep detailed candidate files.
• Corporate teams standardising a hiring process that currently lives in someone’s inbox.
• Agencies that want to grow without hiring administrative staff at the same rate.

Solo recruiter with two calls a week? Probably overkill. Ten or more conversations a day, on a desk with its own object model and its own field structure? That’s the shape this was built for.

Security and privacy

You’re processing sensitive personal data. Names, salaries, personal circumstances shared in confidence, medical limitations, family situations. Not data you hand to any platform that asks.

Two questions, answered separately rather than blurred into one reassuring sentence:

The platform and your candidate data are hosted in the Netherlands, on infrastructure Simply runs. Simply is ISO 27001 certified and the processing agreement is part of the terms.

The AI processing runs inside the EU as well, or on your own key with your own provider. The models themselves are bought in, and they are named in the sub-processor annex to the terms. We say that plainly because it’s the first thing a security review checks: where the processing happens and who built the model are two different questions, and only the first one has a European answer.

Underneath both, permission is checked across four layers on every request: workspace, object, record and field. Authority is re-derived per request rather than frozen at login, field-level PII gating applies to an API response exactly as it applies to a screen, and denied attempts are written to the append-only audit log alongside the allowed ones.

What Simply is not

Honesty works both ways, and this list is on the comparison page too rather than saved for month three.

• No job boards and no multiposting yet. Job board sourcing, multiposting and vacancy distribution are coming; today they are not in the product, so if posting is the centre of your process, this is the wrong system today. Simply does not give you a branded careers site.
• No write-back to your old ATS. A migration is a one-way import. There’s no two-way sync.
• The interface is English. The application doesn’t translate. Documents and CVs generate in Dutch, and Dutch CVs parse fine, but the UI itself is English.
• No SOC 2 and no penetration test. ISO 27001 certified, certificate on request. We have not done a SOC 2 audit and have not commissioned an external pen test.
• No nurture or retention engine. There’s a reminder workflow. There’s no drip campaign engine and no automated re-engagement, because we didn’t build one.
• No autonomous mode. Simply Ask reads your workspace, answers within your permissions and drafts configuration, and every one of those arrives as a proposal a human approves. There is no setting that lets it act on its own.

Each of those is a real reason to choose something else. We’d rather you weigh them now.

Getting started

The cheapest way to find out is fourteen days on the full product, no card. If you want your trial to start with your own vacancy and your own CVs in it rather than an empty workspace, do a half-hour demo first and we’ll set it up that way.

If you’re further along and weighing a switch, the migration page sets out what happens each week, what we need from you, and what your export will and won’t carry across. Bring your current system to the call. We’d rather talk you out of it now than in month four.

For the wider context, read how to get started with AI as a recruiter, see what fully automated conversation processing looks like day to day, or read how AI helps you strengthen candidate relationships rather than weaken them.