Autonomous AI & Pipeline Engineer
Build production RAG pipelines, multi-agent autonomous tool-use workflows, and fine-tuned open-source model integrations for enterprise client automation.
About the role
### Role Overview Nexlyra is pioneering applied enterprise intelligence. We do not build superficial chat wrappers—we engineer robust agentic systems, vector retrieval engines, and multi-step autonomous pipelines that automate complex corporate operations. ### Core Responsibilities - Design and maintain production Retrieval-Augmented Generation (RAG) architectures with hybrid search (dense vector embeddings + sparse BM25). - Build autonomous agent workflows with structured tool calling, deterministic validation loops, and error recovery. - Integrate open-weight and frontier LLM APIs (Groq, Anthropic, OpenAI, DeepSeek) with dynamic failover and token cost optimization. - Implement automated document parsing, token chunking, and semantic vector indexing across institutional knowledge bases. - Benchmark and evaluate model output accuracy, latency, and hallucination rates using automated evaluation rubrics. ### Compensation & Perks - High-growth compensation structure with quarterly impact incentives. - Access to high-throughput compute infrastructure and API credits for experimentation. - Sponsored attendance at top machine learning and software conferences.
Requirements
- 2+ years experience building production AI/ML applications or LLM orchestration workflows. - Expertise in Python and TypeScript, with familiarity with vector stores (pgvector, Pinecone, or Qdrant). - Solid grasp of embedding models, chunking strategies, and re-ranking techniques. - Understanding of prompt engineering, model temperature tuning, structured output generation (JSON schema / Zod). - Strong computer science fundamentals in data structures, concurrency, and API design.