Staff AI Engineer - Production Agents for Fraud & FinCrime

Socratix AI Inc.; San Francisco (CA); Hybrid; USD 180,000 - 260,000; Full time

Benefits offered by this job

Equity

Health coverage (medical, dental, and视

Hybrid schedule – in-office 3 days a‑/

Job summary

Socratix AI Inc. is seeking an early-stage software engineer to design and ship production AI agents that assist fraud and compliance teams.

You will work closely with founders on product strategy, partnering with banks, fintechs, and enterprises to deploy reliable agents that reason about real cases and gather evidence across systems. You’ll help build end-to-end AI agents, establish guardrails and audit trails, and push fast from prototype to production, always focused on solving high-stakes

Qualifications

• 2–4 years of engineering experience or a portfolio showing relevant work.
• Shipped production systems with LLMs or AI agents.
• Strong fundamentals in Python and cloud tooling (AWS) and distributed systems.

Responsibilities

• Build agents end to end: LLM orchestration and multi-agent workflows.
• Own the layer for trust: eval harnesses, guardrails, and audit trails.
• Ship features rapidly with customer contact and real-world fraud contexts.
• Experiment with new models and tooling while keeping what works.
• Turn frontier research into reusable product components.
• Shape roadmap and strategy with founders in a small, fast team.

Skills

LLM orchestration

Python

AWS

Distributed systems

Production systems

Guardrails/audit trails

End-to-end engineering

Tools

MCP integrations

Retrieval systems

LLM tooling

Job description

Socratix AI Inc. is seeking an early-stage software engineer to design and ship production AI agents that assist fraud and compliance teams.

You will work closely with founders on product strategy, partnering with banks, fintechs, and enterprises to deploy reliable agents that reason about real cases and gather evidence across systems. You’ll help build end-to-end AI agents, establish guardrails and audit trails, and push fast from prototype to production, always focused on solving high-stakes