Live Free-Work : Marché IT, ça repart mais quel avenir pour votre job ?

Business-focused summary (market trends & IT job outlook)

1) Market “picking up” — but unevenly

• Startups / scaleups (especially Île-de-France)

  • Recruitment cycles intensified around early September (described as “the war started”) after a calmer August.
  • Hiring is often funding-linked: after fundraising, teams enter a forced-march growth mode where recruiting becomes a priority.
  • There is no universal boom in headcount volume—momentum varies by company type.

• SMEs & large established companies

  • Hiring demand is described as stable vs last year’s September (no clear “boom”).
  • The main change is a shift in job types, not hiring volume.

2) What roles/skills are being hired for (2025–2027 signals)

Top hiring themes mentioned

• AI adoption in production (“industrialization of AI”)
• Infrastructure modernization
• Automation of business processes
• Security is discussed heavily, but participants felt demand was still “lighter” than expected in concrete role openings during that back-to-school wave.

Specific technical hiring patterns

• Full-stack / backend engineers with GenAI and “agent” components
• “FDEs” (deploy(ed) engineer / development-to-client type roles)
  • A mix of coding plus heavy client-facing / solution engineering responsibilities.
  • Often tied to client orgs updating existing software and needing deployment + integration work.
• DevOps / Cloud / CI-CD / CRE-type profiles
  • Needed both for delivery acceleration and cost/ops optimization.
  • Example pressure effect: “merge more PRs per developer per day” under automation/cost constraints.

Frameworks/approach implied (not formally named)

• Skills assessment segmentation
  • HR typically evaluates behavioral fit.
  • Technical AI competence is evaluated later during technical interviews.
• Parallel hiring channel
  • When public ads don’t match actual needs, companies move into a “parallel market” (direct sourcing, agencies, recruiters, pre-scoped staffing).

3) Hiring process realities & “hidden/parallel market”

• Public job ads become less reliable as a signal

  • Many ads run continuously (“phantom ATS offers”), while real needs are handled upstream/downstream.

• How work is staffed in practice (example sequence)

  1. The operational team identifies a need and builds a scope fixed-price or T&M
  2. Candidates are contacted and proposals prepared
  3. After internal steps, buyer/purchasing distributes the request via an MSP / supplier panel
  4. Suppliers are pushed into a formal offer/broadcast process—even if the person may already be selected

• Deal speed & process length

  • For permanent roles, process timeened:
    • from ~1 week–10 days to ~2 to 2.5 weeks
  • Temporary roles were not affected in the same way.

• Actionable access recommendation (to the “parallel market”)

  • Build a shortlist of ~10 relevant target companies
  • Do periodic outreach via LinkedIn:
    • find HR contacts (“People → HR”)
    • send short value-based follow-ups
  • For smaller companies: contact founders or the future manager
  • Goal: become known so you’re considered quickly when openings arise.

4) Daily rates (TJM/TGm) — not uniformly falling

Key points

• The panel emphasized you must understand why rates change:
  • Supply/demand dynamics by profile type
  • Some AI-related skill profiles see rate uplift, while others may compress
• Example uplift:
  • +€50 to +€100/day for certain data/AI-tool-adopting profiles (vs prior periods)

Freelancer “machine” (how missions are routed) affects rates

• When missions pass through referenced IT services companies / MSP panels / procurement-negotiated price lists:
  • purchasing influences pricing
  • freelancers may have less room to negotiate
• In contrast, direct/less constrained channels can support higher TJM

5) “2027 job” — what the next skill set looks like

Named skills/competencies emphasized

• AI engineering + GenAI
• RAG / chains (rag chain) and retrieval-oriented architectures
• Agentic systems
  • what agents are, orchestration, and how they work
• AI product thinking
  • an “AI product manager” must understand business + user sides to make AI products relevant and effective
• Architecting/orchestrating over raw coding
  • expect fewer “raw coding” tasks and more system architecture and orchestration
• Robust system design & maintainability
  • need strong “grand design” sensitivity
  • concern about too many low-value PRs that become hard to maintain

Explicit stance

• “Prompt engineer” is not expected to be a long-term standalone job
  • instead, prompting is treated as a skill set that all AI/product/data/engineering roles should incorporate.

6) Cybersecurity discussion — demand exists but is not always dominant publicly

• The public narrative may overstate where security budget goes; reality can differ.
• Many French contexts have few internal SOCs; work is often outsourced to specialist providers/consultancies.
• The sector is described as growing, with strong opportunity for French talent:
  • driven by reliance on external-country services.
• Hiring evidence (anecdotally):
  • freelance entry via proof-of-skill (e.g., bug bounty / responsible disclosure)
• Certifications vs degrees
  • cyber certifications (including pen test) are valued and seen as “still good” on CVs
  • demonstrated capability is often seen as more important than academic credentials (e.g., confidence from scoring on a CV)

7) Data roles: where AI changes evaluation and demand

• Data teams feel pressure as AI begins to outperform traditional BI in some executive workflows (e.g., a CEO using ChatGPT to challenge BI outputs).
• Hiring pattern mentioned:
  • more senior roles than junior
  • preference for proven pipeline experience (what pipelines were built previously)

8) Remote vs hybrid — more constraints, not just a “market vibe”

• Decision-makers are described as favoring return to office / hybrid.
• For consultants:
  • fully remote is possible but increasingly restricted
    • more common with high-seniority, high-autonomy, and technically demanding roles
    • companies struggle to standardize fairness/management expectations across HR/CFO/CEO alignment
• Common remote transition pattern:
  • start on-site for ~2 weeks, then move hybrid
  • after time (e.g., ~6 months), some teams move to fully remote if performance/governance is proven

---

Key KPIs / metrics / targets mentioned (business execution signals)

• Hiring process cycle time (permanent roles): ~1 week–10 days → 2 to 2.5 weeks
• Rate movement example: +€50 to +€100/day for some AI-enabled data analyst/data-adjacent profiles
• Mandate volume comparison: September 2025 vs September 2026 described as similar/stable
• Market access planning: recommendation to maintain an outreach shortlist of ~10 target companies
• Live audience (contextual, not a business KPI): “several hundred connected… soon 400” (not treated as a business metric)

---

Concrete actionable recommendations extracted

To find jobs in the “parallel market”

• Maintain a small target list of ~10 companies
• Do regular LinkedIn touchpoints with HR decision-makers
• For smaller firms, contact founders or the future manager

To stay competitive for 2027 roles

• Build capability in:
  • agent orchestration
  • RAG / AI chains
  • AI product/business alignment
• Strengthen:
  • system architecture and maintainability
  • (not just coding)
• Treat prompting as a universal skill, not a career title

To defend/future-proof freelance rates

• Understand whether your pipeline routes through MSPs / procurement price lists
• Target channels where you can negotiate more, or where your value is directly measurable

---

Presenters / sources mentioned

• Valérien de Venance (business partner at Suite)
• Pierre (Managing Director at TCO Luxembourg)
• Gaël (associated with FreeWork context; appears as speaker on umbrella/market/remote topics)
• Damien (host; introduced guests and managed transitions/questions)
• FreeWork / WeWork (platform/context referenced; “FreeWork” emphasized)
• Suite, TCO Luxembourg, Allégria (training referenced; owned by Francis Lelon, former founder of Sarenza)