Key skills
AWSAzureCloudDockerGoogle Cloud PlatformJavaScriptKubernetesMicroservicesNext.jsPythonTypeScriptC#C.NETAILLMOpenAIPineconeWeaviateGCPGoogle CloudBedrockGitGitHubRepositorySaaSAtlassianJiraCI/CD
About this role
Role Overview
- Design, develop, and deploy production-grade, secure AI systems utilizing scalable polyglot microservices.
- Integrate state-of-the-art enterprise LLMs (Anthropic Claude Opus 4.6, OpenAI GPT-5.2, Google Gemini 3) into SaaS platforms via dynamic, fault-tolerant API routing gateways.
- Modernize extensive codebases using agentic swarm coding, automated refactoring tools, and cross-language translation.
- Develop autonomous multi-agent workflows utilizing parallel orchestration tools like the OpenAI Codex and Claude Code CLI.
- Automate comprehensive technical documentation for legacy and newly generated codebases using AI orchestration.
- Rapidly prototype → validate via adversarial testing → deploy → iterate.
- Apply good knowledge of software engineering design patterns and architecture to ensure the scalability and maintainability of AI-generated systems.
- Implement complex RAG systems and continually optimize trade-offs between massive-context hydration (up to 1-million tokens) and multi-stage semantic retrieval.
- Work extensively with vector databases (Pinecone, Weaviate, FAISS) and manage persistent document embedding queues.
- Build polyglot microservices specifically capable of handling massive, long-running operational tasks and streaming continuous token generation via persistent connection protocols (WebSockets and Server-Sent Events).
- Ensure SOC2 compliance, data residency (via inference_geo parameters), and deterministic execution through policy-as-code agentic governance.
- Deploy and scale models strictly within secure managed cloud boundaries (Azure AI, AWS Bedrock, GCP Vertex) leveraging Docker and Kubernetes orchestration.
- Architect "zero-touch deployment" CI/CD pipelines fortified with adversarial gating.
- Design and implement unit test and benchmark automation workflows, utilizing AI agentic coding to rigorously validate and test produced code.
- Natively integrate synthetic red teaming into CI/CD to actively prevent prompt drift, reward hacking, and logic degradation.
- Work directly with non-technical stakeholders to translate ambiguous business problems into secure, scalable AI solutions.
- Optimize Atlassian workflows (Jira) to automate the translation of unstructured product requirements into structured prompt contexts for AI agents.
Requirements
- 3-5+ years of demonstrable enterprise software development experience.
- Good knowledge of software engineering design patterns and architecture.
- Flexible in most popular programming languages (Python, Javascript, Typescript, .Net, C#) and web frameworks (NextJS, etc.).
- Ability to switch environments and programming using agentic programming tools to validate and test your produced code.
- Backend development experience designing polyglot APIs, decoupled asynchronous microservices, and utilizing persistent connection protocols (WebSockets/SSE).
- Familiarity with containerization (Docker, Kubernetes) and architecting multi-region, fault-tolerant cloud infrastructure (AWS, Azure, or GCP).
- Enterprise LLM integration and dynamic API routing (OpenAI, Anthropic, Gemini, DeepSeek).
- Multi-agent orchestration and advanced CLI tooling (Claude Code with Auto Hot-Reload and Forking Context, OpenAI Codex).
- Mastery of localized AI IDE layers (Cursor for repository-wide reasoning, GitHub Copilot).
- Massive-context RAG pipelines, dynamic chunking strategies, and vector databases (Pinecone, Weaviate, FAISS).
- Custom evaluations (LangSmith), structured outputs, and managed fine-tuning in secure cloud boundaries.
- Zero-touch CI/CD pipelines and advanced Git workflows (including Git worktree isolation for autonomous sub-agents).
- Unit test and benchmark automation frameworks fully integrated with AI-driven testing.
- Adversarial LLM testing and automated synthetic red-teaming integration.
- Monitoring dynamic access limits and real-time atomic token/credit consumption tracking.
- Strict hallucination mitigation, enterprise guardrails, and deterministic policy-as-code execution.
Tech Stack
- AWS
- Azure
- Cloud
- Docker
- Google Cloud Platform
- JavaScript
- Kubernetes
- Microservices
- Next.js
- Python
- TypeScript