Staff AI Engineer

Socket.dev; New York (NY); On-site; USD 180,000 - 260,000; Full time

Job summary

Obin is seeking AI Engineers to join as one of its earliest hires in New York. You’ll work directly with the founders (ex-Google) and own critical architectural decisions from day one.

You’ll design and deploy LLM-powered agents, collaborate with customers, and own production infrastructure. You’ll balance performance, cost, and safety while shaping core platform design and engineering culture at scale.

Qualifications

• 8-12+ years of software engineering experience.
• 2-3+ years building AI/ML systems, LLM-based agents, or RAG in production.
• Logging, evaluating, optimizing AI applications.
• Sets architecture for agentic AI systems across products; solves non-deterministic challenges.
• Expert in agent-to-agent orchestration, self-correcting agent loops, prompt-routing middleware; mentors senior engineers.
• Proficiency in Python and cloud infrastructure (GCP preferred).
• Experience with cloud infra and APIs.
• Hunger to move fast, own outcomes, and build something enduring.
• Strong intuition around system architecture, performance, and scaling.
• Clear communication in ambiguous, high-stakes problem spaces.

Responsibilities

• Designing and deploying LLM-powered agents using prompt design and orchestration.
• Working directly with customers to co-develop AI solutions tied to ROI.
• Owning production infrastructure and APIs with high availability and observability.
• Evaluating and optimizing systems balancing accuracy, latency, cost, and explainability.
• Contributing to core platform design and reusable deployment infrastructure.
• Setting engineering culture, hiring bar, and long-term technical roadmap.

Skills

Python

LLM frameworks

Agent orchestration

System design

MLOps

Performance optimization

Production tooling

Cloud infrastructure

Tools

GCP

PydanticAI

LlamaIndex

Claude Agent SDK

Job description

The Role:

We're looking for AI Engineers who can help us build and scale intelligent agents from the ground up. You’ll join our team as one of our earliest hires. You'll work directly with the founders (ex-Google) and own key technical decisions from day one.

You’ll be

• Designing and Deploying LLM-Powered AgentsBuild agentic systems using LLMs and RAG to automate complex credit workflows. This includes prompt design, tool use, orchestration logic, and system evaluation.
• Working Directly with CustomersPartner with technical and non-technical stakeholders at major credit firms to understand workflow pain points and co-develop AI solutions that directly tie to ROI.
• Owning Production InfrastructureArchitect and maintain backend systems and APIs that connect to data sources and customer platforms. Ensure high availability, performance, and observability.
• Evaluating and Optimizing SystemsBalance accuracy, latency, reliability, cost, and explainability in our agentic systems. Lead rigorous experimentation and improvement cycles.
• Contributing to Core Platform DesignDefine technical architecture, build reusable infrastructure for agent deployment, and shape the foundation of Obin’s AI engineering best practices.
• Setting Culture and StandardsHelp define Obin’s engineering culture, hiring bar, and long-term technical roadmap

We believe great AI Engineers come from diverse backgrounds and are unified by deep curiosity, pragmatism, and engineering excellence.

Experience Required:

• 8-12+ years of software engineering experience
• 2-3+ years of hands-on experience building AI/ML systems , LLM-based agents and/or RAG systems in production
• Logging, evaluating, optimizing AI applications
• Sets architecture for agentic AI systems across products; solves ambiguous problems inherent to non-deterministic systems (context pollution, runtime state handoffs, safety).
• Expert in agent-to-agent (A2A) orchestration protocols, self-correcting agent loops, prompt-routing middleware; drives AI safety/reliability practices and evaluation standards; mentors senior engineers.
• Proficiency in Python and experience with modern AI/ML frameworks and cloud infrastructure (GCP preferred)
• Experience with cloud infra (preferably GCP) and APIs
• Hunger to move fast, own outcomes, and build something enduring
• Strong intuition around system architecture, performance, and scaling
• Clear communication, especially in ambiguous, high-stakes problem spaces

Bonus:

• Experience with human-in-the-loop systems, feedback loops, and long-horizon agentic task execution
• Background in financial systems, risk modeling, or decision automation
• Familiarity with PydanticAI, LlamaIndex, Google ADK, Claude Agent SDK, or similar frameworks
• You’ve built or contributed to an internal AI platform or reusable LLM tooling