AI Engineer
Agentic Ai
Ai Agent
Ai Agent Platform
Ai Platform
Ai Security
Artificial Intelligence
Artificial Intelligence Engineer
AWS
Aws Bedrock
Aws Sagemaker
Cloud Operations
Data Analysis
Data Processing
Generative AI
Generative Ai Engineering
Information Technology (IT)
Large Language Models
Llm Orchestration
Machine Learning
Programming
Programming Language
Programming Languages
Rag Architectures
Security
Software Development
Software Engineering
Job Description
Socure is hiring a Staff AI Engineer in New York, NY (onsite) to design and ship agentic workflows and internal tools for AI-driven identity solutions.
Responsibilities
- Partner with business teams including Revenue Ops, Marketing, Finance Ops, and Talent Acquisition to catalog manual, high-frequency workflows and rank opportunities by impact, feasibility, and urgency
- Build multi-step, tool-augmented agent workflows that plan, execute, observe outcomes, and iteratively improve
- Design and implement planner–executor and reflection-based architectures to improve reasoning quality and task reliability
- Develop stateful agent systems and implement human-in-the-loop controls such as approval gates, fallback paths, and escalation mechanisms
- Use Socure’s Agentic AI Foundations platform capabilities across agent runtime, orchestration, memory, tool registry, guardrails, and observability
- Provide feedback that informs Agentic AI Foundations platform priorities
- Use LLMs, multi-agent frameworks, and orchestration platforms to build internal solutions that cannot be purchased off the shelf
- Monitor emerging AI technologies and regulatory frameworks to support secure, compliant, intelligent internal systems
- Identify reusable agent design patterns while shipping real workflows, and contribute patterns back to Agentic AI Foundations paved paths
- Apply the evaluation harness and tracing substrate from Agentic AI Foundations for continuous assessment, failure-mode analysis, and tuning speed and accuracy
Requirements
- 8+ years of software engineering experience, including at least 2 years focused on AI/ML systems or LLM-powered applications in production
- Hands-on experience with LLM APIs and at least one agentic framework such as LangGraph, CrewAI, or AutoGen
- Strong Python skills and experience building production-grade backend services, APIs, and data pipelines
- Proven ability to operate in ambiguity: entering undefined problem spaces, defining what to build, and measuring results
- Experience shipping AI/automation solutions that directly impact business operations
- Strong systems thinking focused on reliability, observability, and maintainability from the start
- Ability to work directly with non-technical business stakeholders to understand real workflows and translate them into technical solutions
Technologies
- Python
- LLM APIs, LLMs
- LangGraph, CrewAI, AutoGen
- Multi-agent frameworks, orchestration platforms
- Agentic AI Foundations, planner–executor, reflection-based architectures
- Evaluation harness, tracing substrate
- AWS-hosted LLM infrastructure: Bedrock, AgentCore/Strands, Lambda, SageMaker
- Agent tooling: LangSmith, Weights & Biases, Arize
- RAG pipelines, vector databases, knowledge graphs
- Memory/grounding systems
- MCP (Model Context Protocol), function calling
Preferred
- Production experience with multi-agent systems, workflow orchestration, or complex tool-use patterns
- Familiarity with evaluation frameworks for agent reliability, accuracy benchmarking, and failure-mode analysis (LangSmith, Weights & Biases, Arize, or custom eval pipelines)
- Experience with RAG pipelines, vector databases, knowledge graphs, or memory/grounding systems
- Experience with real-time or streaming AI systems
- Familiarity with AI safety and security practices including prompt injection prevention, hallucinations mitigation, and data protection/privacy
- Background in a regulated industry such as fintech, healthcare, or government
- Experience with agent skill abstraction and structured tool integration via MCP (Model Context Protocol), function calling, or similar protocols
- Experience with AWS-hosted LLM infrastructure (Bedrock, AgentCore/Strands, Lambda, SageMaker) or equivalent cloud ML services
Compensation
- $220,000 - $290,000 per year