AI LLM Technology Architecture Manager

YOU ARE

As an experienced and senior AI/LLM  Architect, you will play a pivotal role in designing and delivering end-to-end AI platform architectures that power the modern, reinvented enterprise. Operating at the intersection of business and engineering, you will own the technical design of advanced AI systems — spanning classical machine learning, generative AI, and agentic systems — ensuring they are purposefully architected to meet client business objectives and enterprise-grade standards.   

Within this scope, you will take deep ownership of one or more critical architecture domains — such as agentic application design, AI security and trust, AI operations and observability, data and knowledge engineering, or model platforms and inference — serving as the lead authority in your domain across client engagements. You will develop and maintain specialized expertise in the technologies, patterns, and emerging practices within your domain, bringing that depth to bear in shaping architecture decisions, accelerating delivery, and building reusable assets that extend across the practice.  You will evaluate, select, and apply the right design patterns, technical frameworks, and tools within your domain and across the broader AI architecture — ensuring cohesion across the full system. This includes architecting AI agents encompassing multi-agent orchestration, tool use, skills use, and memory systems, as well as the integration of fine-tuned foundation models and classical ML models into scalable, production-ready platforms.   

A critical dimension of this role is owning the design of a comprehensive AI context layer — drawing on enterprise knowledge bases, structured and unstructured data sources, and domain-specific content — to ground AI systems in the realities of each client's business and ensure outputs are accurate, trustworthy, and impactful.    As the technical authority on your domains, you will lead architecture decisions and be accountable for ensuring systems meet rigorous non-functional requirements across security, observability, governance, performance, and scalability. You will produce and own the architecture artifacts that shape delivery — including architecture decision records (ADRs), component and data flow diagrams, and integration specifications — and provide the technical leadership that enables cross-functional teams of data engineers, ML engineers, and application developers to execute with clarity and confidence.   

Your contributions will be instrumental in shaping how clients adopt and scale AI, pushing the boundaries of what these systems can achieve and delivering measurable, lasting business value.

THE WORK

• Translate business strategy into a technical vision by defining the non-functional requirements (NFRs) necessary to meet operational goals for performance, reliability, and cost.
• Lead stakeholder workshops to align on technical feasibility, define project scope, and manage expectations with clients and leadership.
• Drive the technology selection process, evaluating build-vs-buy decisions for AI platforms (e.g., Arize, LangSmith) and foundational models.
• Architect model- and tool-agnostic multi-agent systems governed by an MCP Control Plane.
• Design and implement the Agent Registry as the mandatory system of record and the AI Gateway for runtime policy enforcement.
• Design and implement a certification gate to ensure no uncertified agents enter production, validating identity, policies, and evaluation metrics.
• Design, implement, and abstract core agent services, including a first-class abstracted memory service with semantic, episodic, and procedural endpoints.
• Architect the end-to-end data pipeline for AI systems, including data ingestion, preprocessing, and synchronization for fine-tuning and RAG.
• Design and implement the context layer—spanning knowledge graphs, vector search, and semantic retrieval—to create reliable, grounded RAG pipelines.
• Architect foundation model adaptation strategies, including dynamic, cost-and-performance-aware model routing and selection.
• Design, implement, and prototype high-throughput, low-latency inferencing solutions using techniques like response caching and request batching.
• Define security, governance, and observability as centrally-enforced, by-design controls for all AI systems.
• Architect a robust, defense-in-depth security framework, including per-agent identity with IAM/IAP binding and layered guardrails.
• Design and implement FinOps controls enforced at the AI Gateway, including token budgets, cost-center labeling, and threshold alerts.
• Establish the framework for comprehensive system evaluation, adopting productized tools and instrumenting observability with OTel
• Define and maintain the enterprise-wide AI reference architecture, reusable design patterns, and a library of approved software components.
• Independently design, implement, build, and deliver proof-of-concept prototypes and foundational software components to validate architectural decisions.
• Produce and own authoritative architecture artifacts, including blueprints, sequence diagrams, design specifications, and Architectural Decision Records (ADRs).
• Mentor and guide cross-functional engineering teams (data, ML, application) on architectural best practices and design patterns.
• Continuously research and integrate emerging AI patterns, frameworks, and technologies to maintain a forward-looking architecture.