Job Summary:
BMG360 is a high growth, private equity backed, tech enabled marketing agency that specializes in direct response advertisements. They are seeking an Agentic Solutions Engineer to design, develop, and maintain API-first infrastructure that supports autonomous marketing workflows and AI-driven client solutions.
Responsibilities:
• Design, build, and maintain RESTful and GraphQL APIs that serve as the backbone of agentic workflow systems.
• Implement Model Context Protocol (MCP) servers that expose tools, resources, and prompts to LLM agents—enabling Claude and other agents to interface with internal and external data and services.
• Develop webhook receivers, event-driven triggers, and real-time communication layers (WebSockets, SSE) to support autonomous agent pipelines.
• Ensure APIs are secure, versioned, rate-limited, and observable with proper logging and distributed tracing.
• Build and maintain the internal MCP server registry; keep tool schemas, descriptions, and input validation accurate and up to date.
• Build and extend Django-based backend services that power internal tools, client data pipelines, and agent orchestration APIs.
• Design PostgreSQL schemas for agentic workflow state, job queuing, and audit trails.
• Implement Django REST Framework (DRF) endpoints consumed by React frontends and external agent clients.
• Develop background task processing using Celery and Temporal or similar for async agent jobs, polling loops, and scheduled automations.
• Write Python integrations with LLM providers (Anthropic, AWS), marketing platform APIs (Meta, Google, GA4), and data warehouses (Snowflake).
• Build internal admin dashboards using React-Admin to surface agentic workflow status, inspect agent job runs, and manage MCP server configurations.
• Develop React-based monitoring and configuration UIs for the Architect and delivery teams.
• Create reusable React components and hooks for common UI patterns across internal tool surfaces.
• Integrate frontend tooling with Django/DRF backends and MCP-driven data sources.
• Develop and maintain a reusable Python + TypeScript SDK of pre-built agent tools—platform API wrappers, data warehouse clients, webhook handlers—so new MCP servers and workflows can be stood up faster.
• Support the Architect in prototyping and evaluating new agentic frameworks (Flowise, LangChain, custom orchestration).
• Document MCP servers, API contracts, and SDK usage so the broader engineering team can build on your infrastructure.
• Test and validate the accuracy and reliability of API outputs, MCP tool responses, and agentic workflow integrations.
Qualifications:
Required:
• 3–6 years in backend or full-stack engineering roles, ideally at an agency, SaaS, or data/AI company.
• Expert-level Python: you write clean, tested, idiomatic Python for APIs, data pipelines, and LLM integrations daily.
• Proficient TypeScript/JavaScript: comfortable in both Node.js server contexts and React frontend work.
• Hands-on experience building and consuming REST APIs—you understand OAuth2, JWT, API versioning, pagination, and webhook security.
• React experience: you can build functional, component-driven UIs and are familiar with hooks, state management, and API integration patterns.
• Familiarity with relational databases (PostgreSQL preferred) and ability to write efficient queries and design schemas.
• Understanding of async patterns: event queues, background workers, webhooks, polling.
• Python: FastAPI, Django/DRF, Celery, Pydantic, httpx, pytest—production-grade Python is your default.
• TypeScript/JS: Node.js APIs, React (hooks, context, functional components), type-safe interfaces, module patterns.
• Django: models, migrations, DRF serializers, viewsets, permissions, admin customization, Celery task integration.
• React-Admin: custom resources, data providers, list/show/edit views, auth integration—or strong React with a fast ramp.
• APIs: OpenAPI spec-first design, Swagger/Redoc documentation, rate limiting, CORS, versioning strategies.
• MCP Protocol: familiarity with Model Context Protocol preferred; hands-on MCP server development is a major plus.
• Cloud/infrastructure: AWS or GCP (Lambda, Cloud Run, S3/GCS), Docker, basic CI/CD pipelines (GitHub Actions).
• Data: PostgreSQL, Snowflake familiarity; ETL patterns; ability to write complex SQL.
Preferred:
• Familiarity with Model Context Protocol preferred; hands-on MCP server development is a major plus.
Company:
BMG360 BMG360 is a technology-enabled marketing service company. Founded in 2003, the company is headquartered in Shelton, USA, with a team of 51-200 employees. The company is currently Growth Stage.