Senior AI Developer
Location: New York, United States
Type: Full-time
Department: Technology
Job Summary
We are seeking a Senior AI Developer to support one of our premier clients—a leading global financial institution—with strong expertise in building intelligent AI agents and components that can reason, plan, and act autonomously. The ideal candidate will have hands-on experience developing scalable multi-agent AI systems using modern orchestration frameworks such as LangChain and LangGraph, integrating agentic workflows end-to-end, and shipping production-grade AI applications.
Responsibilities
- Design and develop AI agents and autonomous multi-agent systems using modern agentic frameworks including LangGraph and LangChain, with the ability to architect agent graphs, define node transitions, and manage stateful agent workflows
- Build and orchestrate multi-agent pipelines—including supervisor agents, collaborative agent networks, and hierarchical agent architectures—to solve complex, multi-step financial use cases
- Implement guardrails, reasoning workflows, and ReAct-based patterns within LangChain/LangGraph to improve reliability, decision-making, and agent safety
- Develop memory management (short-term, long-term, episodic) and tool-use capabilities (MCP, LangChain Tools, custom tool integrations) for AI agent systems
- Leverage LangGraph's stateful graph execution model to build resilient, interruptible, and human-in-the-loop agentic workflows
- Integrate LLM-powered agents with external APIs, databases, and enterprise data platforms via LangChain's retrieval, routing, and chain composition primitives
- Partner closely with prompt engineers, data scientists, and platform teams to optimize AI application performance across multi-agent deployments
- Build and maintain scalable Python-based services, APIs, and microservices that serve as agent execution environments and tool backends
- Develop and support AIOps capabilities and CI/CD pipelines for AI agent deployment, versioning, and monitoring (including LangSmith or equivalent observability tooling)
- Work with modern data platforms including Snowflake, Databricks, and Lakehouse architectures as grounding and tool-use data sources for agents
- Ensure AI agent solutions are scalable, secure, observable, and production-ready
Eligibility Requirements
- 8+ years of overall software engineering experience, with a strong focus on AI/ML systems in recent years
- Hands-on production experience with LangChain — including chains, agents, tools, retrievers, memory modules, and prompt templates
- Hands-on production experience with LangGraph — including stateful graph construction, conditional edges, checkpointing, human-in-the-loop interrupts, and multi-agent graph topologies
- Demonstrated experience designing and deploying multi-agent systems — including orchestrator/worker patterns, agent-to-agent communication, task delegation, and shared state management
- Experience implementing guardrails, ReAct patterns, and chain-of-thought reasoning within agentic pipelines
- Strong understanding of agent memory architectures (in-context, vector-store-backed, episodic) and tool-use patterns (function calling, MCP, LangChain tool wrappers)
- Familiarity with LangSmith or equivalent observability/tracing platforms for debugging and monitoring agent behaviour in production
- Strong Python engineering skills including async programming, APIs, and microservices
- Experience with AIOps and CI/CD pipeline development for AI agent deployment and lifecycle management
- Hands-on experience with Snowflake, Databricks, and Lakehouse architectures
- Strong understanding of scalable distributed systems and cloud-native application development
- Strong communication and cross-functional collaboration skills
- Nice to Have
- Experience with other agentic frameworks such as AutoGen, CrewAI, or OpenAI Assistants API
- Familiarity with LangGraph Cloud or self-hosted LangGraph Server for agent deployment
- Background in financial services AI applications (risk, compliance, trading, operations)
- Experience with vector databases (Pinecone, Weaviate, pgvector) as long-term memory stores for agents
- Contributions to open-source LangChain/LangGraph ecosystem
In the US, the target base salary for this role is $125,000-$140,000. Compensation is based on a range of factors that include relevant experience, knowledge, skills, other job-related qualifications, and geography. We expect the majority of candidates who are offered roles at our company to fall throughout the range based on these factors
How to Apply
- Click "Apply Now" to submit your resume through our career site
- Be sure to include any relevant experience that aligns with the role.
- Qualified candidates will be contacted by a member of our recruitment team for next steps
About eClerx
eClerx is a leading provider of productized services, bringing together people, technology and domain expertise to amplify business results.
The firm provides business process management, automation, and analytics services to a number of Fortune 2000 enterprises, including some of the world’s leading financial services, communications, retail, fashion, media & entertainment, manufacturing, travel & leisure, and technology companies. Incorporated in 2000, eClerx is traded on both the Bombay and National Stock Exchanges of India. The firm employs more than 19,000 people across Australia, Canada, France, Germany, Switzerland, Egypt. India, Italy, Netherlands, Peru, Philippines, Singapore, Thailand, the UK, and the USA.
For more information, visit www.eclerx.com
You can also find us on:
https://www.linkedin.com/company/eclerx/
https://www.indeed.com/cmp/Eclerx/about
https://www.glassdoor.com/eClerx
eClerx is an Equal Opportunity Employer. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, national origin, disability or protected veteran status, or any other legally protected basis, in accordance with applicable law. We are also committed to protecting and safeguarding your personal data. Please find our policy here
