Principal Architect- Agentic AI (Deep-Tech R&D)

ON
On-site
Posted Oct 9, 2026
1 position

Opens LinkedIn

Employment type: Full-time
Experience level: Lead · 10+ years
Posting language: English
Working hours: 40 hours per week
Seniority: Mid-Senior level
Application method: Direct apply is available

Job summary

Drive agentic AI research and development, establish reference architectures and evaluation frameworks, and own the autonomy roadmap. Transfer deep-tech intellectual property to the Center of Excellence and delivery teams, and maintain the technical peer relationship with Anthropic.

Job details

Role Purpose The deep-tech authority and the BU's autonomy engine — drives novel agentic-AI R&D, reference architectures and the IP moat that compounds toward autonomous delivery. $$ Target — Revenue & Value Mandate Reusable agents / reference architectures / eval frameworks created; IP reuse value; technical moat. Monetised through accelerator reuse across the portfolio. Core Responsibilities •Drive novel agentic-AI R&D and the deep-tech agenda •Create reference architectures and eval frameworks •Own the autonomy roadmap •Hand deep-tech IP to CoE and the pods •Hold the Anthropic technical peer relationship Ownership Boundaries — Owns •Deep-tech R&D agenda, reference architectures and eval frameworks •The autonomy roadmap Ownership Boundaries — Does Not Own •Delivery throughput (FDE / CoE) •Account P&L (LoB) Required Capabilities •Distinguished-level agentic-AI / LLM R&D •Eval / guardrails depth •Frontier-AI peer credibility Preferred Capabilities •Published research / patents •Anthropic-stack mastery

What you’ll do

Requirements

Requires distinguished-level expertise in agentic AI and large language model research and development, deep knowledge of evaluation and guardrails, and credibility with frontier AI peers. Published research or patents and mastery of the Anthropic stack are preferred.