Production AI Systems Engineer - Client-Embedded

Elios AI; New York (NY); On-site; USD 150,000 - 210,000; Full time

Job summary

Elios AI is hiring Forward Deployed Engineers to architect, deploy, and operate AI systems in real-world client environments. You’ll translate ambiguous business problems into production-grade architectures integrating LLMs, data pipelines, and operational workflows.

You’ll design multi-step agent workflows, build scalable backend services, and collaborate with clients to iterate rapidly with real-user feedback, ensuring systems perform in production.

Qualifications

• 3–8+ years of software engineering experience.
• Strong proficiency in Python and/or TypeScript.
• Experience building distributed systems, APIs, and data‑intensive applications.

Responsibilities

• Design and deploy LLM-powered applications using APIs (OpenAI, Anthropic, open‑source models).
• Build agent-based systems using orchestration frameworks (LangChain, LangGraph, custom agents).
• Develop evaluation pipelines for non-deterministic outputs (prompt testing, scoring, guardrails).
• Build and optimize data pipelines (batch + streaming) using tools like Airflow, Kafka, or similar.
• Design data models and integrate structured/unstructured data sources (APIs, warehouses, document stores).
• Work with data platforms such as Snowflake, Databricks, BigQuery, or Postgres.
• Implement caching, indexing, and retrieval strategies for performance and cost optimization.
• Build scalable backend services and APIs (Python, FastAPI, Node.js, TypeScript).
• Deploy services using Docker, Kubernetes, and cloud platforms (AWS, GCP, Azure).
• Implement CI/CD pipelines and versioning for models, prompts, and workflows.

Skills

Python

TypeScript

Distributed systems

APIs

Tools

Airflow

Kafka

Snowflake

Databricks

BigQuery

Postgres

Job description

Elios AI is hiring Forward Deployed Engineers to architect, deploy, and operate AI systems in real-world client environments. You’ll translate ambiguous business problems into production-grade architectures integrating LLMs, data pipelines, and operational workflows.

You’ll design multi-step agent workflows, build scalable backend services, and collaborate with clients to iterate rapidly with real-user feedback, ensuring systems perform in production.