Running recruitment without a separate ATS: hiring inside your HR system

Why a separate ATS is overkill under 200 people

Dedicated applicant-tracking systems were built for companies that hire constantly: a recruiting team, dozens of open roles at once, agencies to coordinate. Under 200 people, hiring doesn't look like that — it comes in bursts. Three roles after a new client lands, then nothing for a quarter.

Run a standalone ATS through that reality and you pay three times:

• Tool sprawl. Another subscription, another login, another admin surface — billed year-round for hiring that happens in bursts.
• Double data entry. Candidates live in the ATS, employees live in the HR system, and every hire gets re-typed between them — name, contacts, salary, start date. Two records for one person drift apart from day one.
• Handoff loss. The ATS's job ends at "hired". Start date, equipment, accounts, the first-week checklist — all of it falls into email between two systems, which is exactly where onboarding steps get lost.

The fix isn't a better ATS. It's hiring living in the same system the hire lands in.

What a hiring flow actually needs at this size

Strip away the enterprise recruiting machinery and a 5–200-person company needs five things:

1. A place to define and publish roles — a job description and a careers page you don't need a developer to update.
2. One pipeline view — every candidate, every stage, visible at a glance. A kanban board, not a spreadsheet with color-coding conventions only one person remembers.
3. Structured feedback — a scorecard per interview, so decisions rest on written assessments instead of hallway impressions that evaporate within a week.
4. GDPR hygiene — consent, retention and deletion handled by the system, not by someone remembering to purge a CV folder.
5. A real handoff — "hired" should start onboarding, not end recruiting.

What you don't need below 200 people: requisition approval chains, agency portals, recruiter workload dashboards, multi-brand career sites. That machinery is what enterprise ATS pricing pays for — and what sits unused when you buy it too early.

The flow, part one: from open role to applications

1. Define the job. Title, seniority, location or remote, and the salary range if you publish one — increasingly expected, and EU pay-transparency rules are pushing the market that way. The job description is the classic blank-page stall; this is one place AI drafting genuinely helps, as long as a human edits before anything goes public.

2. Publish to your careers page. A lightweight public page under your company's name: open roles, an apply form, done. Share the link on LinkedIn, in communities, with your network — no separate website project, no webmaster in the loop for every edit.

3. Let applications land in one place. Every application should create a candidate profile with the CV attached. Parsing the CV into structured fields — contacts, skills, experience — kills the re-typing, but keep the original document: parsing is an accelerator, not the record.

That alone fixes the most common small-company failure: candidates arriving through email, LinkedIn DMs and referrals into three inboxes, with nobody certain who was replied to and who was silently dropped.

The flow, part two: from screening to day one

4. Screen with AI assistance, decide with humans. When fifty applications arrive for one role, ranking them against the job description is real time saved. But the ranking must stay a suggestion — a person reads, decides and moves each candidate. No auto-rejection, ever. That's better hiring, and it matters legally: the GDPR restricts fully automated decisions that significantly affect people, and a rejection is exactly that.

5. Run interviews on a kanban. Stages that match your real process — screen, technical interview, final, offer — with candidates dragged between them and a scorecard filled in per interview. When two finalists sit side by side, you compare written assessments, not memories.

6. Make the offer. Record what was offered and when, and let the candidate accept online instead of playing scanned-PDF ping-pong.

7. Hand off, don't drop off. The hire becomes an employee record in the same system — nothing re-typed — and their onboarding starts immediately. The stretch between "signed" and "productive" is where new hires form their first impression of you; a structured first week is the cheapest retention tool you'll ever deploy.

GDPR: candidate data is personal data

Candidates have the same data-protection rights as employees, and recruiting data is messier — CVs carry addresses, photos, sometimes family details. The working rules:

• Consent and transparency. Tell applicants what you store and why, at the point of application — not in a policy nobody can find.
• Retention with an end date. Decide how long you keep candidates after a rejection, and get consent if you keep them for future roles — then make the cleanup automatic. A CV folder from three hiring cycles ago is a liability, not a talent pool.
• Data rights. A candidate can ask what you hold and ask you to delete it. Answering must take minutes, not an archaeology dig through mailboxes and shared drives.
• No silent automation. If software scores people, a human must own every decision — see the screening rule above.

None of this is legal advice: if you hire at volume or across borders, have your counsel or DPO walk the flow once. But a flow where consent, retention and deletion are system features — not calendar reminders — is the difference between compliance you have and compliance you claim.

How Helia HR does this

The Recruitment pack turns hiring on inside Helia — $99/month flat for the whole company, unlimited users (see pricing), so the founder, two team leads and an office manager can all work the pipeline without per-seat math:

• Jobs with AI-drafted descriptions — a solid draft in seconds; you edit, you publish.
• A public careers page for your company, live the moment a job is published.
• CV upload with AI parsing into structured candidate profiles — the original file stays attached.
• AI screening and ranking, always human-gated — the AI suggests, a person decides. Nobody is rejected by a machine.
• Interview kanban and scorecards — your stages, drag-and-drop, written feedback per interview.
• Duplicate detection — repeat applicants are flagged so the history stays on one profile.
• GDPR built in — consent captured at application, a retention window with automatic cleanup, deletion on request.
• One-click candidate → employee handoff — the hire becomes an employee record and the department's onboarding checklist is created automatically.
• Recruiting analytics — time-to-hire and pipeline conversion per stage, so you see where candidates stall.
• Redeployment board — before hiring externally, see who's on the bench and could fill the role. For a services company, one caught redeployment is worth more than the pack costs.

It all rides on the same roles, permissions and audit log as the rest of Helia — getting started covers the wider setup.

FAQ

When does a dedicated ATS actually make sense?

When recruiting is a full-time function: several recruiters, always-on hiring, agency coordination, multi-brand career sites. If you employ a recruiting team, buy recruiting-team software. Below 200 people that's rarely the shape of the problem.

Do we pay per recruiter seat?

No — the Recruitment pack is $99/month flat per company, unlimited users. Everyone who interviews can leave a scorecard without anyone counting licenses.

Can the AI reject candidates automatically?

No, deliberately. AI drafts job descriptions, parses CVs and ranks applicants — every decision that affects a candidate is made by a person. Automated rejection is bad hiring and a GDPR minefield; Helia doesn't offer it.

What happens to candidates we didn't hire?

They're kept for the retention window you set, with consent recorded, and cleaned up automatically when it lapses. A candidate who requests deletion sooner can be removed on the spot.

We already track candidates in a Trello board or a spreadsheet. Why move?

The board isn't the problem — the isolation is. No consent trail, no retention rules, CVs scattered in personal drives, and every hire re-typed into the HR system. Moving the same kanban inside the HR system keeps the workflow you like and fixes the compliance and the handoff you're missing.