Back to Jobs
Recently postedOther

Technical Lead, AI Platform

DynPro Inc.United States🇺🇸United States

Quick overview

SeniorityMid Senior
Work modeHybrid
LocationUnited States
DockerFastAPISQLGitAWSOAuthGitHub ActionsGoogle CloudGrafanaHelmJWTJavaJenkinsKafkaLLMPrometheusPythonRedisTerraformVaultVercelgRPCGenerative AIRESTAzureReactPostgreSQLKubernetesCI/CDLangChainLangGraphRAG

Job description

We're building an enterprise platform for AI agents. As the Technical Lead, you'll lead development of the platform end to end, guide a team of engineers, and stay hands-on in the code.
You'll work hands-on across every part of the platform:
  • Agent harness and SDK: the frameworks and tools engineers use to build and run agents.
  • AI infrastructure and governance: secure, policy-controlled access to models, tools, and data.
  • Context and memory: getting agents the right information at the right time.
  • Evaluation and observability: measuring agent quality, cost, and behavior before and after release.
  • AI user experiences: how agents appear in applications, assistants, and partner integrations.
  • Applied agents: building production agents on the platform alongside domain teams.

What you'll do

  • Own the design, and delivery of the platform, from first design through production operation.
  • Write code, review designs and pull requests, and take on the hardest problems.
  • Build reusable platform capabilities that other teams can depend on, with clear interfaces, versioning, and documentation.
  • Set and uphold engineering standards for quality, security, observability, and evaluation.
  • Improve latency, reliability, and cost for production AI workloads.
  • Run what you build: define SLOs, take part in on-call, and lead incident reviews.
  • Partner with architects, product managers, security, and domain teams.

Minimum qualifications

  • Bachelor's degree in computer science or a related field, or equivalent practical experience.
  • 8+ years building and running production backend or distributed systems, including 2+ years as a technical lead.
  • Hands-on experience shipping LLM-powered applications or agents to production, and handling failure.
  • Strong Python , and experience designing frameworks, SDKs, or platform APIs that other engineers build on.
  • Solid distributed systems fundamentals: API design, asynchronous and event-driven processing, streaming, caching, rate limiting, and fault tolerance.
  • Clear written communication, including design docs that explain trade-offs to both engineers and leadership.

Technical skills

We're looking for experience in most of these and depth in several:
  • Agent development: tool and function calling, multi-step and multi-agent orchestration, task delegation, durable execution, and human-in-the-loop approvals.
  • Agent interoperability: exposing agents and tools to other applications and agents through open protocols.
  • Context engineering: retrieval-augmented generation, vector and hybrid search, short- and long-term memory, summarization, and tool selection across large tool catalogs.
  • LLM evaluation: eval datasets and rubrics, LLM-as-judge, regression testing in CI, and online evaluation of production traffic.
  • LLM infrastructure: routing across model providers, token and cost management, response caching, and latency optimization.
  • AI safety and governance: guardrails, prompt-injection defense, PII detection and redaction, policy engines, and sandboxed code execution.
  • Security: authentication and authorization, service-to-service identity, least-privilege access, tenant isolation, and audit logging in multi-tenant systems.
  • Observability: distributed tracing, metrics, logging, SLOs, and LLM observability.
  • AI user experiences: streaming agent output to web clients, and generative or dynamic UI.
  • Cloud-native infrastructure: containers, container orchestration, CI/CD, and infrastructure as code.

Languages, frameworks, and technologies

Python is a must. For the rest, we're looking for hands-on experience across most of these categories.
  • Languages: expert Python; working knowledge of SQL. Java is a plus.
  • Agent frameworks: LangGraph and LangChain, or comparable frameworks such as the OpenAI Agents SDK, Google ADK, or Pydantic AI.
  • Agent protocols: MCP, A2A, A2UI.
  • Models and providers: APIs from major model providers and cloud AI platforms, such as OpenAI, Anthropic, Google Gemini, Amazon Bedrock, and Azure OpenAI; serving open-weight models with vLLM; LLM gateways such as LiteLLM.
  • Backend and APIs: FastAPI, asyncio, and Pydantic; REST, gRPC, Server-Sent Events, and WebSockets; event streaming with Kafka or similar.
  • Data and retrieval: PostgreSQL with pgvector; search and vector engines such as OpenSearch, Elasticsearch, or Qdrant; Redis; embedding and reranking models.
  • Evaluation and observability: OpenTelemetry, including its GenAI semantic conventions; LLM tracing and eval platforms such as LangSmith; eval libraries such as Ragas or DeepEval; Prometheus and Grafana.
  • Security and governance: OAuth 2.0, OIDC, and JWT; workload identity such as SPIFFE/SPIRE; policy as code with Open Policy Agent or Cedar; secrets management such as HashiCorp Vault; guardrail and PII tooling such as NeMo Guardrails, Llama Guard, or Presidio.
  • Cloud and infrastructure: Docker, Kubernetes, and Helm; Terraform; at least one major cloud (AWS, Azure, or Google Cloud Platform); CI/CD such as GitHub Actions or Jenkins; sandboxed execution with gVisor or Firecracker.
  • Front end: React; streaming chat and agent interfaces; generative UI toolkits and specs such as the Vercel AI SDK, A2UI, or MCP-UI.

Preferred qualifications

  • Experience building an internal platform, SDK, or framework that multiple engineering teams adopted.
  • Experience running AI systems for external enterprise customers.
  • Contributions to open-source agent frameworks or protocols.

Similar jobs