evidoria

← Back to browse

Good practice Imported

ChatFT and Match FT — France Travail's AI Tools for Jobseeker Matching and Advisor Support

France · Paris · See the France profile · See the Paris profile

Evidence: Descriptive / self-reported Top 25% 73/100 · Ask Evidence Copilot about this practice

France's public employment service rolled out ChatFT to nearly 40,000 advisors from November 2024, saving up to 3 hours/week each, while a 6-region Match FT pilot lifted training enrolment by 17% and cut recruitment time by over 3 days, ahead of a 2026–2027 national rollout.

up to 3 hours
Advisor time saved per week (ChatFT) (from November 2024)
+17%
Training enrolment increase (Match FT Formation pilot) (pilot, 6 regions)
-3 days (average)
Recruitment time-to-fill reduction (Match FT Offres pilot) (pilot, 6 regions)
~40,000 advisors (>50% of workforce)
ChatFT advisor reach (from November 2024)
62%
Jobseekers reporting positive impact from AI tools (survey published 23 January 2025)

Details

Maturity
Scaling
Promoter
France Travail (the French national public employment service, formerly Pôle Emploi)
Period
November 2024 (ChatFT rollout) – 2026–2027 (Match FT national rollout)
Keywords
employment services, generative AI, job matching, public administration

Context

France Travail, France's national public employment service (formerly Pole Emploi), built a family of AI tools on models from French AI company Mistral to support both its advisors and jobseekers. ChatFT, an internal generative-AI assistant for drafting job offers, summarising interviews and writing emails, reached nearly 40,000 advisors — more than half of France Travail's workforce — from November 2024. A meeting-transcription extension, ChatFT Ecoute, was piloted in six regions ahead of a planned 2027 rollout. Separately, Match FT is a matching system with two tracks piloted in six regions: Match FT Offres matches jobseekers to open positions, and Match FT Formation matches jobseekers to training programmes. National rollout is scheduled for end of 2026 (Match FT Offres) and early 2027 (Match FT Formation).

Results

In pilot regions, France Travail reported that ChatFT saved advisors up to three hours a week, with a stated ambition to redirect the equivalent of 800 full-time positions into more face-to-face jobseeker support. Match FT Offres was associated with an average reduction of more than three days in time-to-fill in its pilot regions, and Match FT Formation with a 17% increase in training enrolment. A France Travail-published survey (23 January 2025) found 77% of jobseekers had used AI tools in their job search, 62% reported a positive impact, and usage was higher among under-25s (83%) than over-50s (69%). These figures are self-reported by France Travail via its own infographic and public statements, corroborated by independent press (franceinfo, LeMagIT) but not yet by an independent academic or auditor evaluation, and national-scale rollout has not yet occurred.

Implementation

Indicative cost
High (€500k–€5M) — No budget figure is published; classified as high cost given workforce-wide (~40,000 advisor) deployment across a national public employment service, built on a proprietary LLM provider.
Time to results
Medium (1–3 years) — Phased rollout from ChatFT's November 2024 launch to Match FT's dated national rollout in end-2026/early-2027, roughly 2.5 years.
Staffing & skills
France Travail advisors (~40,000 reached via ChatFT), Director of AI Projects (named responsible official)

Conditions for success

  • Built on a French/EU sovereign AI provider (Mistral)
  • Multi-region piloting (6 regions) before national rollout
  • Defined, dated national rollout milestones (end-2026 / early-2027)

Common failure modes

  • Figures are self-reported by France Travail with no independent academic or auditor evaluation to date
  • Advisor unions and commentators have raised unquantified concerns about workload and de-skilling as AI tools expand

Commonly funded by

National / regional programmes ESF+ — European Social Fund Plus

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful