Senior DevOps

Weflayr SAS; Paris; Hybrid; EUR 80,000 - 120,000; Full time

Benefits offered by this job

Equity

Job summary

Weflayr SAS, Paris-based AI gateway startup, seeks an Infrastructure & Security lead to own the platform resilience and LLM inference stack on-site in Paris. You will drive scalable, secure systems while collaborating closely with the CTO and tech team.

You will migrate workloads to Kubernetes, improve production efficiency, and maintain SOC 2 / ISO 27001 readiness. Equity may be part of the compensation package, reflecting the early-stage environment.

Qualifications

• 5+ years of professional experience building infrastructure at scale.
• Deep experience with at least one hyperscaler.
• Proficiency in Terraform and Kubernetes.
• Strong security foundations.
• Experience deploying LLMs in production.
• Code with AI and plan/validate work.

Responsibilities

• Migrate ECS/Lambda stack to Kubernetes if relevant.
• Improve production stack in a cost-efficient way.
• Help maintain SOC 2 / ISO 27001 certifications.
• Serve open-source LLMs distributed by our gateway.
• Training / serving of fine-tuned models (LoRA).

Skills

Kubernetes

Terraform

Hyperscaler

Security

LLM in prod

Tools

vLLM

DVC

GitHub Actions

Job description

Weflayr is an AI gateway built to give its users the highest level of control and quality on dynamic model routing. It routes using each user’s own past data: each candidate model is benchmarked on it, and the routing engine uses the results to pick the best model for each call.

We raised a $2.5M pre-seed and are incubated in the F/AI program. We have opened the product to a small group of beta testers and plan to release it globally in the coming weeks.

About the role

You will own the infrastructure & security of Weflayr as we scale, making sure it remains resilient at high throughput. You will also build & maintain our LLM inference stack. Example tasks that you will be doing:

• Migrate our ECS/Lambda stack to Kubernetes if relevant
• Improve our production stack as we scale in a cost-efficient way
• Help maintain our SOC 2 / ISO 27001 certifications
• Serve open-source LLMs distributed by our gateway
• Training / serving of fine-tuned models (LoRA) (PS: it may come a bit later on)

We are a very horizontal team: you will work hand in hand with our CTO, Noé Achache, and the rest of the tech team.

On site in Paris: we truly believe that to achieve great things, we need to be together, to celebrate the good times, support each other through the harder ones, and above all because, at such an early stage, the most important decisions are made in informal discussions. We do offer some flexibility to suit everyone’s needs, e.g., working remotely one day a week, or a full week a couple of times a year.

What we are looking for

You are a good fit for this job if:

• You are looking for a big challenge, and want to take on a high level of ownership, with the level of proactivity and dedication that is needed to succeed. You want to take pride in contributing to the success of Weflayr as a whole, rather than succeeding on your own individual subjects.
• You have 5+ years of professional experience building and running infrastructures by tens of thousands or more
• You have deep experience with:
  • at least one hyperscaler (not necessarily AWS)
  • Terraform
  • Kubernetes
• You have strong foundations in security
• You have already served an LLM in production with all the run that comes with it
• You write 100% of your code with AI, but still use your engineering skills to plan/validate your work
• You have very good soft skills (communication, adaptability, collaboration, …)
• Although you are a technician, you also care deeply about the product, and you take pride in seeing your work being used by real users.
• Being pragmatic is your core value: what drives you is to build things that work when needed rather than building something super advanced that may not be needed.

Technical stack

• Others: vLLM, DVC, GitHub Actions

Our team

We are a small team (currently 3 people + freelancers) working hand in hand.

Our founders:

• Romain Mazeries, co-founder & ceo: Former co-founder and CEO of MangoPay, a payment fintech company he built and scaled up to 120M€ in annual revenue.
• Noé Achache, co-founder & cto: Former Head of AI Engineering at Theodo Data & AI, where he led the development of various ML/DL projects and AI agents in production.

Compensation

80k€ to 120k€+ and equity depending on profile.

Process

• 45 min engineering skills with Noé (CTO)
• 30 min with Romain (CEO)
• Coffee or lunch with the founders, to make sure there is a mutual fit

Due to the large number of applications we receive, we won’t be able to reply to candidates who are not based in Paris, as all jobs are on site .

Benchmark your real calls, approve the models you trust, and put them to work with a routing strategy built around your business.