SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Vibe.co is an audience-first streaming TV advertising platform acquired by Walmart in August 2026. The company enables marketers to reach targeted audiences, optimize campaigns in real time, and measure business outcomes with precision comparable to digital advertising platforms. Vibe serves 12,000+ brands reaching 120+ million households.
The Performance team owns advertiser outcomes: prediction models, bidding engine, pacing, and delivery infrastructure. This role reports to Clément Schiano, Engineering Manager for Performance.
You will own reliability across the production stack (prediction serving, bidding engine, pacing, delivery). Key responsibilities include:
- Define SLOs and build observability to track them; shape on-call practices
- Reduce p99 latency on the serving path and optimize bidding engine replica counts without sacrificing latency
- Work on the data layer for real-time serving, focusing on Redis access patterns, caching strategy, and degradation behavior
- Build end-to-end stress testing in staging to identify platform limits before production encounters them
- Handle capacity planning and infrastructure cost management as advertiser volume grows
- Collaborate closely with data scientists shipping models and with Benoît Delsart (platform/tooling engineer)
- Own incident response and ensure postmortems drive shipped improvements
- Collaborate with other engineering teams, especially DSP, on system interactions
The role exists because infrastructure work has split into two jobs: getting data scientists from experiment to production quickly, and keeping the production serving path fast, cheap, and reliable. Currently one person handles both; this role takes over the reliability half. Scale is the core challenge—Vibe was acquired by Walmart and plans to run many more advertisers through the same bidding stack. Every bid is a real-time decision with advertiser budget behind it; p99 latency movements directly impact advertiser delivery and company revenue.
The Performance team is seven people, giving you room to set direction on reliability. The stack already exists and works; you're not starting from a blank page.
Vibe operates on three values: Impact (tying goals to real business outcomes), Ambition (thinking 10x), and Urgency (moving daily). The company offers variable pay based on objectives, hybrid flexibility (3x/week in Paris office), health insurance via Alan, meal vouchers via Swile, annual team offsites, and quarterly in-person tech syncs.
REQUIREMENTS:
- Owned latency-sensitive services in production with on-call responsibility; experience being paged for serious incidents
- Solid Kubernetes and cloud fundamentals; comfortable debugging p99 latency problems
- Observability experience: metrics, tracing, logs, and judgment about alert thresholds
- Infrastructure as code and CI/CD; ability to ship fixes, not just identify them
- Comfortable owning systems you didn't design; will be on-call before understanding every model in the stack
NICE TO HAVES:
- High-QPS, low-latency serving experience in ad tech, trading, gaming, streaming, or similar
- Deep debugging experience with Redis, Kafka, or similar systems
- Capacity planning and cost optimization on large infrastructure footprints
- Experience with model rollouts and how they differ from service rollouts
- Background in programmatic advertising, auctions, or CTV
- Paris-based (company supports relocation/visa sponsorship)
- Experience with Go, Rust, or Python
