Member of Technical Staff – Identity / Agent Identity

Colossus Technologies Group; United States; On-site; USD 150,000 - 230,000; Full time

Job summary

Colossus Technologies Group, a fast-growing cybersecurity startup, is seeking an identity engineer to build the infrastructure that enables autonomous AI agents to securely authenticate and act across enterprise systems.

You will design and implement identity platforms, OAuth/OIDC flows, and fine-grained policy engines, ensuring every action is auditable and authorized. This role emphasizes engineering depth over administration.

Qualifications

• Experience building identity platforms, authentication or authorization systems.
• Experience with OAuth/OIDC infrastructure.
• Knowledge of delegated authorization and policy-based access control.
• Strong software engineering fundamentals with a focus on identity protocols and credentials.

Responsibilities

• Build identity infrastructure for autonomous AI agents with credentials, permissions and lifecycle.
• Design authentication and authorization for agents operating across enterprise systems.
• Develop delegated authorization mechanisms so agents can act on behalf of users without over-privilege.
• Implement OAuth/OIDC and machine-to-machine authentication flows for agentic systems and long-running workflows.
• Create fine-grained authorization and policy systems evaluating users, agents, tools, resources and context.

Skills

Identity engineering

OAuth/OIDC

RBAC/ABAC

Machine-to-machine authentication

Policy engines

Distributed authorization

Job description

Member of Technical Staff – Identity / Agent Identity

Colossus Technologies Group is working with a fast-growing, early-stage cybersecurity company building the security and identity layer for enterprise AI agents, MCP servers, skills and plugins.

Traditional identity systems were built around humans, applications and service accounts. AI agents create a different problem — they can operate autonomously, act on behalf of users, access multiple tools and systems, and continue taking actions without a human actively involved.

We're looking for an identity engineer who wants to build the infrastructure to solve that problem.

The team is working on engineering problems around:

• Building identity infrastructure that gives autonomous AI agents their own identities, credentials, permissions and lifecycle
• Designing authentication and authorization for agents that need to operate independently across enterprise systems
• Building delegated authorization that allows an agent to act on behalf of a user without simply inheriting everything that user can access
• Designing OAuth/OIDC and machine-to-machine authentication flows for agentic systems and long-running autonomous workflows
• Building fine-grained authorization and policy systems that evaluate users, agents, tools, resources and runtime context before allowing an action
• Solving credential and token management for agents that may interact with dozens of APIs, MCP servers and enterprise applications
• Building systems that can determine exactly who initiated an action, which agent executed it, what credentials were used and why the action was authorized
• Designing identity infrastructure that can scale from individual AI assistants to enterprises running thousands of autonomous agents

We're interested in engineers who have built identity platforms, authentication or authorization systems, OAuth/OIDC infrastructure, policy engines, delegated authorization, RBAC/ABAC systems, credential platforms or other large-scale IAM infrastructure.

Strong software engineering ability is important. This is not an IAM administration or implementation role. We're looking for someone who has built the underlying systems and understands identity protocols, tokens, credentials and authorization at the engineering level.

Especially interesting backgrounds include identity infrastructure, developer identity, cloud IAM, Zero Trust, policy-as-code, workload identity, machine identity, secrets/credential management and large-scale distributed authorization systems.

The ideal person understands identity deeply enough to know where the existing models start to break when software can reason, use tools and take actions autonomously — and wants to build what comes next.