Senior AI Engineer, Tools & Agents

About the role

TwelveLabs is building the intelligence layer for video, the dominant medium of information on the planet, by creating multimodal AI models that understand content across sight, sound, and motion. The Senior AI Engineer, Tools & Agents will own the critical integration layer that makes this complex technology accessible and usable for developers and AI agents in production environments. You will be responsible for designing and operating the primary surfaces through which external systems interact with TwelveLabs' video understanding capabilities. This role sits at the intersection of advanced AI systems and reliable software engineering, requiring equal parts research insight and operational pragmatism. You will translate the capabilities of the underlying Jockey system into robust, trustworthy, and scalable interfaces for the broader ecosystem. The work you do will directly determine how enterprises and developers build on top of TwelveLabs' video AI platform. You are expected to bring a builder's mindset to every problem, balancing technical excellence with speed and clarity. Your contributions will shape the foundation upon which customer applications and agent workflows are built.

Key facts

What you'll do

Design and operate the agent-facing integration surface, including TwelveLabs' MCP server and agent-to-agent interfaces, defining how AI agents discover and leverage video understanding capabilities.

Architect the auth and access substrate for the platform, implementing OAuth, RBAC, and multi-tenant isolation from the ground up to meet enterprise security requirements.

Build the enterprise-readiness layer, including per-API-key usage tracking, rate limiting based on real system constraints, and reliability engineering to ensure platform stability under heavy load.

Own the end-to-end experience of the MCP server, covering agent discovery, tool invocation, failure mode handling, and the evolution of the agentic ecosystem interface.

Establish the research partnership loop, working closely with the research team to turn frontier video and image understanding models into durable, production-grade product surfaces.

Define the contracts and versioning strategies for external developers, ensuring that the integration surface is both powerful and sustainable over time.

Operate as the primary owner of the developer experience for external and internal users, making decisions that balance flexibility with consistency and clarity.

Drive the platform behavior under pressure, setting up the monitoring, throttling, and reliability mechanisms that protect enterprise customers from instability.

Collaborate across product, research, and engineering teams to align the integration layer with the broader vision for video AI understanding and application.

Lead the design of supporting infrastructure such as metering, observability, and logging to ensure transparency and trust in the platform.

Make tradeoffs across the full stack of the tool and agent platform, from retrieval and reasoning primitives to authentication and rate limiting.

Champion best practices in software engineering for AI systems, ensuring that the surfaces you build are maintainable, testable, and scalable.

Translate complex system behaviors into clear documentation and interfaces that enable other teams to build effectively on top of the platform.

Ensure that the integration layer supports both rapid experimentation in research and the stability required for production deployments.

Contribute to the architectural vision for how agents and humans will interact with video intelligence in the future.

Requirements

You have shipped real agentic or LLM-powered systems in production, not demos, and you can discuss in detail the failure modes you encountered and how you resolved them.

You understand how retrieval, reasoning, and agent orchestration work together as a unified system, and you design integration surfaces that reflect this reality under real conditions.

You have owned a developer-facing or external API end-to-end, making real decisions about contracts, versioning, and the experience of building on your surface.

You have built authentication and authorization in production, with direct ownership of OAuth, OIDC, RBAC, or multi-tenant isolation, rather than relying solely on third-party frameworks.

You have thought seriously about enterprise-readiness, including the gap between a working capability and a trusted platform that large organizations will adopt.

You bring versatility across the full platform surface and are comfortable owning the entire stack of the tool and agent integration layer.

You have operated in fast-moving environments such as startups, hypergrowth companies, or AI-first teams, shipping features under ambiguity without heavy process overhead.

You are comfortable making technical decisions with incomplete information and iterating based on feedback from users and internal stakeholders.

You communicate clearly in writing and speaking, able to align technical constraints with product goals and customer needs.

Nice to have

Experience with MCP servers and agent tooling is preferred.

Background in building or consuming video or image understanding systems is beneficial.

Familiarity with production AI infrastructure, including serving, monitoring, and observability, is a strong advantage.

Practical notes

Location: San Francisco. Onsite or hybrid. No fully remote option.

Employment is full-time.