AI-native HR · v3.0› shipped May 7

HR is the work,
not the forms.

Lerio is the AI HR platform for remote-first teams. Agents that hire, review, pay, and protect — drafted from Slack, Linear, GitHub, and the jurisdiction rules you already have to follow. Humans approve where judgement is required.

14 agentsin production
178jurisdictions modelled
91%draft acceptance
< 60stime to first action
leriomanager briefingagent ready
AG
Performance agent· just now
Yes — and softly. Maria's first-reply latency on Slack threads dropped 40% in 7 days, while her PR throughput is unchanged. That pattern usually precedes disengagement, not workload — try a low-stakes 1:1 this week.
Evidence:#Slack· 10 threads · 7d windowLatency· 23m → 14m medianGitHub· 11 PRs · normal cadence§Policy· Wellbeing Research v2.4
Trusted by remote-first teams
HalcyonPivot LabsNorthwindMercato
1,240 companies€84M payroll / wkSOC2 II · ISO 27001

Four roles. Four cold-opens. No empty prompt.

Every Lerio surface starts with what matters today, not a navigation tree. The agent has already read your week. You either approve, ask a follow-up, or override.

EM
Elena · Senior DesignerLisbon · Mon 9:14 · monday-morning briefing
EMPLOYEE

Your week: 3 meetings, 2 PRs in review. You haven't taken PTO in 94 days — August has good team coverage if you want to book something. Also: your right-to-work doc expires in 6 weeks. Want me to handle the renewal?

Suggested:

Not features. Coworkers.

Each agent owns a domain end-to-end — capable of drafting, citing its sources, and explaining its reasoning. Pass-rates below are from our public eval harness, run weekly.

OBOnboarding94.1%

Provisions environments, drafts 30/60/90 plans from the job description and the team's actual repos, books intros from calendar metadata.

access & equipment30/60/90 draftsintrosbuddy match
See an evidence example
PFPerformance89.7%

Drafts reviews from PR descriptions, Linear scope, Slack collaboration metadata. Flags scope mismatch and stagnation. Never reads message content.

review draftsscope match1:1 promptsgrowth plans
See an evidence example
CEComp & Equity92.4%

Benchmarks against live market data per role + level + jurisdiction. Drafts merit changes, equity refreshes, comp letters with diffs and soft-confirm.

benchmarksmerit cyclesequity refreshletters
See an evidence example
CPCompliance96.8%

Models 178 jurisdictions and the actual citations. Watches contractor day-counts, document expiries, PE-risk thresholds. Surfaces the rule, not the panic.

PE riskdoc renewalscontractsmulti-juris payroll
See an evidence example
WBWellbeing83.2%·β

Detects burnout and flight-risk signals from collaboration patterns — not from message content. Suggests check-ins. Manager-only surface. Beta.

flight-riskPTO nudgesburnout signalconsent-gated
See an evidence example
+In trials

Recruiting (live demos with 12 design partners), Procurement, Learning & Development. Closed beta — request access if you'd like to shape one.

recruitingL&Dprocurement
Closed beta →

Every claim ships with an evidence drawer.

Click any number, any verdict, any draft — the right rail expands with source PRs, Slack metadata, jurisdiction rules with citations, comp benchmark data. Opacity kills the product.

Performance agent · claim

“Maria's response latency dropped 40% this week — might be worth a check-in.”

generated 11:42 · model v3.0.4confidence 0.87
Show evidence — 4 sources, 1 citation
Evidence4 sources · trailing 7d
Slack · response latency 2 sources
#
Median first-reply time on threads where Maria is @mentioned10 threads · 7d window · vs. prior 30d baseline
slack/api
Drop from 23m14m median first-reply (−39.1%)statistically significant at p < 0.05 over 14d window
analysis
GitHub · review patterns 1 source
PR review acknowledgement time up 2.4× over same window11 PRs reviewed · #design-system · #portal-rewrite
gh/pr/4821
Policy · the recommendation reason 1 citation
“Sustained drops in collaboration responsiveness — independent of output volume — predict disengagement at 68% within 6 weeks.”— Lerio Wellbeing Research, internal eval set v2.4 · open paper ↗

A jurisdiction-aware answer, every time.

Pick a contractor scenario. Lerio's compliance agent surfaces PE-risk thresholds, tax posture, and the contract template — with the actual citation. This is the live agent, not a screenshot.

Compliance · scenario builder

Is this contractor safe to keep?

Configure where they live, where they work today, and how long they've been there. The agent applies live jurisdiction rules and returns a verdict + sources.

Or press
CautionAt 178 days in Portugal, approaching the 183-day PE-risk threshold. Consider: limit further days, or move to an EOR in Portugal.
Tax regime
MEI / Pessoa Jurídica
PE threshold
183 days · OECD Art. 5
Contract template
BR-CTR-PJ
Days remaining (PE)
5 days
Citation: OECD Model Tax Convention, Art. 5 §1–3 · Portugal Brazil bilateral treaty. The 183-day rule is the safe-harbor; PE may attach earlier with a fixed place of business.— Lerio jurisdiction model · last revised 2026-04-22 · open the full rule ↗

Market data, sourced and shown.

Pick a role + market. The comp agent returns a defensible range with the sources behind it — not a single number you'd have to defend in a hallway.

Comp & Equity · benchmark

What's market for this hire?

RangeBase salary band for this role & market, adjusted for company stage. P50 is the typical close.
€68k€84kbase · annual
Market
Lisbon · PT
Stage multiplier
1.00×
Equity guidance
0.06 – 0.14%
Sample size
n = 236
Sources: Levels.fyi (n=143, trailing 6mo) · Figures.hr EU-tech panel (n=240) · your last 3 closes · OECD wage indices. Auto-refreshed weekly.— Lerio Comp model v3.0 · last refresh 2026-05-08 09:00 UTC

We grade our own agents in the open.

Every Friday, Lerio re-runs 2,847 labeled scenarios across the agent suite and publishes the pass-rates here. Failures get filed. Regressions get rolled back automatically.

AgentPass-rate · 7dΔ vs. priorScenariosConfidence interval
Onboarding94.1% +0.6412[92.1 – 95.4]
Performance89.7% +1.2624[88.1 – 91.0]
Comp & Equity92.4% +0.3518[90.9 – 93.7]
Compliance96.8% +0.4894[95.8 – 97.5]
Wellbeing · β83.2% −2.1399[81.2 – 85.0]

Last run · Fri May 8, 2026 · 09:00 UTC · full results & methodology ↗

Changelog.

Plain language. Dated. Public. Full log ↗

May 8, 2026Shipped
Receipts get an Undo countdown. Every agent action now lands as a persistent receipt with a 30-second Undo. Sensitive actions (comp, terminations, contracts) require soft-confirm before commit.v3.0.4 · Performance, Comp & Equity, Admin surfaces
May 6, 2026Eval
Compliance agent +1.4% on the multi-jurisdiction subset. New scenarios added for Spain's nómada digital visa and Brazil's MEI day-count edge cases.methodology · 178 jurisdictions covered, 2,847 scenarios in harness
May 2, 2026Fix
Performance agent stopped citing GitHub squash merges twice. A double-count bug inflated PR throughput for repos with a squash-and-merge default. Reviews drafted before May 2 have been re-flagged for re-review.root cause · 218 reviews re-flagged · all affected managers notified
Apr 28, 2026Shipped
Founder cold-open. A new role briefing tuned to burn, headcount, jurisdiction expansion, and equity. Replaces the legacy "Reports" tab — no migration needed.v3.0.0 · Founder surface · GA

Per-employee. All agents included.

No "AI add-on" tier. No "compliance add-on". The agents are the product. Pay one number, ship from day one.

Starter
€18/ employee / month
Up to 25 employees. All five agents. Slack & GitHub integrations. Two jurisdictions.
  • Onboarding, Performance, Comp, Compliance, Wellbeing
  • Evidence drawer & eval harness
  • Email + Slack support
Most installed
Team
€32/ employee / month
25–250 employees. Multi-jurisdiction payroll. Linear, Notion, Zoom integrations. Custom briefing tuning.
  • Everything in Starter
  • Multi-jurisdiction payroll (178 modeled)
  • API + webhooks · custom integrations
  • Dedicated CSM · 4-hour response SLA
Scale
Custom/ talk to us
250+ employees. Private eval set. SAML, audit log export, regional data residency.
  • Everything in Team
  • Private agent fine-tuning on your eval set
  • Data residency (EU / US / SG)
  • Annual on-site security review

Bring your team. We'll bring the agents.

30-minute live demo using your real Slack and GitHub. You'll see the manager briefing for one of your actual teams — anonymized — by the end of the call.

No sales tricks. Calendly link sent within an hour — typically by a founder.