Principal AI Platform Architect

Cyclotron, Inc.

About The Position

We are hiring a senior, hands-on engineer to partner directly with the Head of Products in architecting and building a new multi-agent orchestration and communication platform from the ground up. In the early phase, this person will act as a high-trust technical counterpart focused on architecture, core implementation, and platform design. As the product matures, they will become the application owner and help lead a growing offshore team supporting ongoing development.

Requirements

7+ years of software engineering experience building production systems.
Deep experience delivering LLM-based systems or AI-enabled products in production, not just prototypes or demos.
Strong understanding of agent orchestration, including multi-agent coordination, routing, tool use, memory, workflow state, and human-in-the-loop patterns.
Experience with one or more agent ecosystems such as Vercel AI SDK, LangChain, LangGraph, OpenAI Agents SDK, or similar frameworks.
Strong understanding of MCP-style integration patterns, tool contracts, and external-system interoperability.
Experience with vector databases, embeddings, retrieval pipelines, and semantic memory strategies.
Strong experience designing and operating cloud-native services on Azure.
Solid software engineering fundamentals across architecture, debugging, reliability, and scaling.
Experience working in GitHub Enterprise environments.
Prior experience acting as a technical lead, system owner, staff engineer, or principal engineer on a meaningful product or platform.

Nice To Haves

Experience with graph-based orchestration and graph databases.
Experience with Convex or other real-time application backends.
Experience with mem0.ai, cortexmemory.dev, or similar memory-layer tooling.
Experience integrating AI systems into Microsoft Teams.
Experience with Dynamics 365 / CRM data models, APIs, and workflow patterns.
Experience generating structured business artifacts such as proposals, statements of work, reports, or Word documents.
Experience building configurable template systems for non-technical business users.
Experience with multimodal pipelines involving text, documents, images, and file understanding.
Experience leading distributed and/or offshore teams.

Responsibilities

Partner directly with the Head of Products to define the platform architecture, implementation strategy, and long-term technical direction.
Architect and build a reliable multi-agent orchestration framework for coordinating specialized agents, tool calls, memory systems, and enterprise integrations.
Design clear communication patterns between orchestrator, agents, tools, memory providers, and external systems.
Build the initial platform capabilities, including agent UI, memory integration, multimodal workflows, and enterprise communication channels.
Create a Statement of Work generation capability that can produce structured Word documents from templates, business rules, and configurable user inputs.
Design a templating and administration model that allows non-technical users to customize output structure and business logic without engineering support.
Design a Dynamics 365 agent or integration layer capable of extracting, interpreting, and operationalizing complex CRM context more effectively than conventional copilots.
Establish strong patterns around observability, tracing, evaluation, retries, fallbacks, safety, and reliability for AI-native systems.
Serve as the long-term application owner as the platform grows, ensuring architectural consistency, engineering quality, and delivery discipline.
Provide technical leadership for future supporting engineers, including an offshore development team, through architecture guidance, code reviews, work decomposition, and mentoring.