Senior AI Agentic Systems Architect

Jobtailor

Austin (TX)

On-site

USD 180,000 - 240,000

Full time

14 days+

Get more replies from employers

Send a job-specific resume in minutes.

Job summary

TeamViewer seeks a senior architect to own and evolve the agentic platform across TeamViewer ONE. You will define reference MCP/server architectures, tool catalogs, and context assembly while establishing rigorous evaluation, observability, and reliability standards.

You will mentor engineers, run architecture reviews, and partner with security to design safe, scalable AI-driven systems for enterprise use.

Qualifications

10+ years of software engineering experience.
Demonstrated expertise in complex architectures.
Strong Python and TypeScript skills.
Deep knowledge of distributed systems.
Extensive experience designing and delivering agentic systems, including tool-using agents, MCP or comparable protocols, retrieval, context engineering, multi-agent orchestration, and evaluation frameworks.
Proven track record bringing AI solutions into enterprise production environments.
Strong understanding of security, compliance, multi-tenancy, cost management, reliability, and operational excellence at scale.
Ability to define quality standards, develop evaluation approaches, interpret results, and identify real-world failure modes in AI systems.
Deep understanding of hallucinations, context degradation, prompt injection, non-determinism, silent regressions, and architectural controls to mitigate them.
Regular use of AI coding agents with critical review practices and accountability for correctness, security, maintainability, and architectural alignment.
Ability to communicate complex architectural concepts to technical and non-technical stakeholders.
Ability to balance business value with technical considerations while collaborating across global teams.
Solid knowledge of software licensing, open-source compliance, and AI dependency risk.
Bachelor’s or Master’s degree in Computer Science, Software Engineering, or a related field.

Responsibilities

Own the architecture of TeamViewer’s agentic platform, including Tia, third‑party agents, tool use, retrieval, memory, orchestration, and governed actions across TeamViewer ONE.
Define reference architecture for MCP servers and clients, tool catalogues, context assembly, memory, orchestration, and trust boundaries.
Establish evaluation and observability discipline by defining quality criteria, tracing agent behavior, catching regressions before release, and using results to guide architectural decisions.
Set positions on model choice, routing, fallback behavior, cost governance, vendor dependency, and production reliability.
Design safety architecture with security partners, covering prompt injection containment, permission models, tenancy isolation, auditability, and AI component risk.
Guide teams hands‑on through early project phases.
Conduct architecture and design reviews.
Mentor senior engineers.
Maintain clear architecture documentation for engineers and agents.