Solutions Architect (AI)

Role:

Location:

Type:

Department:

Experience:

About the Role

We are looking for a hands-on Solution Architect & Team Lead who knows that a great AI model

is useless without a great API wrapping it. You won’t just be optimizing prompts; you will be

architecting the distributed systems that serve them.

You will bridge the gap between experimental AI and rigorous software engineering. You will

lead a team to build production-grade, fault-tolerant systems where Clean Code, API standards,

and System Design are just as important as the LLM context window.

Key Responsibilities

•

System & API Architecture:

(REST/gRPC/GraphQL). Define the interface between our AI agents and the frontend/client

applications, ensuring low latency and type safety.

•

Production-Grade Engineering:

unit/integration testing, code reviews, and design patterns (SOLID, DRY). Ensure the codebase

is maintainable and modular, not a "script" thrown into production.

•

Agentic Systems & MCP:

integrations. Implement the Model Context Protocol (MCP) to standardize how models interface

with external data and tools.

•

Retrieval (RAG) at Scale:

re-ranking, treating the vector database as a critical production component, not a sandbox

experiment.

•

Team Leadership:

Drive technical decisions on build vs. buy, schema design, and microservices architecture.

•

Security & Reliability:

Technical Requirements

1. Core Software & API Engineering (Essential)

•

Mastery of Backend Development:

Node.js/TypeScript. You write clean, typed, and self-documenting code.

•

API Design Excellence:

internal comms), and GraphQL. Experience with OpenAPI/Swagger specifications is

non-negotiable.

•

Database Design:

You understand how to optimize queries for high-throughput applications.

•

Distributed Systems:

to handle asynchronous AI tasks and long-running agent processes.

Generative AI & Agents

•

Agentic Workflows:

LangGraph, or AutoGen. You understand state management in multi-turn conversations.

•

Model Context Protocol (MCP):

LLMs read/write context and interact with client-side applications.

•

LLM Integration:

(Claude), and open-source models (Llama 3, Mistral).

3. MLOps & Production AI

•

Retrieval (RAG):

•

Observability:

Datadog) to track latency, token usage, and errors.

•

Evaluation:

4. Cloud & Infrastructure

•

Cloud Native:

•

Containerization:

containers independently of stateful vector stores.

•

Infrastructure as Code:

Success Metrics

•

System Stability:

•

Code Quality:

•

Performance:

strategies.

Why Join Us?

•

Engineering First:

"

component, not a magic box.

We build robust software where AI is a

•

Architectural Freedom:

we use to scale.

•

High Impact: