About this role
Network Solutions seeks a Senior Software Engineer - AI to build the agent-powered back end of its AI platform. The role focuses on scalable, model-agnostic services, multi-agent workflows and retrieval-augmented generation for a platform intended to serve millions of users.
Responsibilities
- Design and scale asynchronous REST and WebSocket APIs using Python 3.11+ and FastAPI, with dependency injection, type hints and vertical-slice architecture.
- Implement handoff, sequential and concurrent workflows in Semantic Kernel to route work among specialised LLM agents. Integrate OpenAI GPT-4.1/mini and Google Gemini 2.5 Flash through a provider-agnostic layer supporting A/B testing and cost-aware routing.
- Deliver RAG using Azure AI Search, pgvector or Chroma. Expose tool-using OpenAI Assistants with Code Interpreter for data analysis and file manipulation.
- Manage database schemas with SQLModel, SQLAlchemy 2 and Alembic, and tune PostgreSQL for high-concurrency asynchronous access. Own CI/CD for linting, type-checking, testing, Docker packaging and deployment using Bitbucket and Jenkins.
- Instrument services with structlog JSON logs, OpenTelemetry traces and cost and latency metrics, targeting p95 latency below 100 ms. Promote AI-assisted development with GitHub Copilot and Cursor.
Requirements
At least 5 years building production APIs in Python and at least 2 years with FastAPI or a similar asynchronous stack. Candidates need deep knowledge of async I/O, Pedantic v2 as stated in the posting, dependency injection and observability; experience with Semantic Kernel or a comparable agent framework; practical RAG implementations; strong PostgreSQL and migration skills; and LLM API integrations or side projects involving structured output. Poetry, virtual-environment isolation, end-to-end CI/CD ownership, analytical ability and problem-solving skills are required. The role requires remote-work readiness and daily overlap of at least 09:00–13:00 EST.
Preferred
Experience with RabbitMQ, Azure Service Bus or Kafka, and with Grafana or Lang Fuse for LLM cost governance. The posting describes HMO benefits of 175k MBL with one free dependent after one year of service, a punctuality bonus, vacation, and learning and development opportunities.