Vercel

vercel.com · 70 open roles in the last 6 months · hiring in 5 countries

Loading…

Vercel appears to be investing in an agent-oriented developer platform, while expanding the infrastructure, enterprise capabilities, and commercial support needed to run it at scale.

Inferred from 85 job postings from the last year, with recent ones weighted more · updated 21 Oct 2026. Not a statement by the company.

What changed recently

The latest month adds a distinct Software Factory product bet: AI agents that handle ongoing software maintenance, alongside a more agent-powered Dashboard. It also sharpens investment in platform fundamentals—storage, compute, networking, and scheduled work—and adds Trust & Safety machine learning; these build on the preceding months’ AI Gateway, agent framework, and agentic infrastructure work.

New bethigh confidenceMostly new this month

Putting agents across the software development lifecycle

1m 33m 46m 61y 6

Vercel is shaping a suite of agent-facing developer experiences, from the new Software Factory area for continuous bug fixes, feature work, performance improvements, and security mitigation to an agent-powered Dashboard that surfaces insights and recommended actions. Roles also point to the underlying agent stack: eve is a framework and runtime built on Workflows, AI SDK, and Sandbox, while agentic infrastructure supports millions of builds.

The Software Factory product manager posting calls it a new product area.
The Dashboard product manager describes a shift from a project listing page to an agent-powered surface.
The eve role describes a framework and runtime for building and operating production AI agents.
Agentic build and deployment infrastructure handles millions of builds.

Scalinghigh confidenceMostly new this month

Scaling compute, storage, networking, and observability

1m 33m 46m 61y 6

Recent product roles span Fluid Compute, Vercel Functions and Sandboxes, networking and CDN, observability, and a storage portfolio including Blob and Drives. The scale and workload details are concrete: the platform handles more than 6 million deployments a day, over 2 million Sandboxes, and a trillion requests per month; the new storage role also calls out scale, compliance, and data locality needs.

Product leadership roles appeared recently across storage, compute, networking/CDN, and observability.
Fluid Compute is microVM-based and supports Sandboxes for untrusted and agent-generated code.
The CDN role cites billions of requests across 20+ regions.
Observability work includes Runtime Logs, OpenTelemetry tracing, and Drains.

Scalinghigh confidence

Building enterprise adoption and delivery capacity

1m 13m 36m 41y 6

A new Professional Services leadership role would establish a global model spanning Technical Account Management and Forward Deployed Engineering, from prototypes through production rollouts. Recent EMEA solutions architecture and deal-desk hiring, alongside commercial sales roles, points to more structured technical support and deal execution for enterprise customers.

The Head of Professional Services role defines the engagement model, delivery methodology, and team.
Solutions Architecture leadership is being built for EMEA and startups.
Deal Desk and commercial legal roles focus on quote-to-contract execution and faster deal processes.
Forward-deployed roles work directly with strategic customers on migrations and production AI systems.

Scalingmedium confidence

Automating platform abuse detection and security response

1m 13m 16m 41y 4

A recent Machine Learning Engineer role centers on production systems to detect and disrupt abuse across Vercel’s platform. Earlier security roles describe automated vulnerability analysis, incident response, cloud controls, and IAM, suggesting security work spans both customer-facing platform trust and internal infrastructure.

The Trust & Safety role explicitly calls for machine-learning systems for abuse detection and classification.
Product Security is exploring LLM- and agent-based vulnerability analysis.
Detection and response work covers servers, containers, and endpoints across a multi-cloud fleet.

Scalinghigh confidence

Making AI model access practical for production workloads

1m 03m 16m 21y 3

AI Gateway connects applications and agents to hundreds of models through one API key, with provider routing and fallbacks, unified billing, and observability. Earlier engineering and developer-relations roles focus on reliability, low latency, and high-volume inference, while a recent design role focuses on model discovery, first requests, and operating workloads in production.

AI Gateway postings describe access to hundreds of models and name OpenAI, Anthropic, and Google.
The engineering role emphasizes reliable, low-latency systems for high-volume inference.
Recent design hiring focuses on the public model catalog and production operations.

Where they are expanding

EMEA, especially London and Berlin — Recent roles add EMEA commercial deal support and solutions-architecture leadership, following hiring in London and Berlin for sales, partnerships, and partner delivery. This looks like continued investment in regional enterprise coverage rather than a clear indication of a newly opened market.

Questions to ask in a coffee chat

How is the Software Factory team deciding which maintenance tasks agents should handle first, and where do developers stay in the loop?
Is the Dashboard’s move toward agent-powered recommendations changing how its product and engineering teams work together?
How do teams coordinate changes across Fluid Compute, Sandboxes, storage, and scheduled work for agent-generated workloads?
What customer problems are shaping the priorities for AI Gateway as the model ecosystem changes?
How are Forward Deployed Engineering and Solutions Architecture turning customer implementation patterns into reusable product improvements?
How does Trust & Safety work with product teams when abuse-detection systems need to balance automation and review?

What this can't tell you

Job postings show areas of hiring and stated responsibilities, but do not establish how much engineering or budget each product receives or whether every described initiative will ship.
Some signals come from individual product or leadership roles, so the detail and maturity of those efforts may be difficult to infer from postings alone.