Lead Software Engineer - AI/ML - AI Agent Platform
Salary not listed • Jersey City, NJ, United States • Full-time
Posted 1mo ago
No salary is listed. Many US states require employers to post a pay range, so ask about pay early.
Job Location
Jersey City, NJ, United States
Tech Stack
Remote Work Policy
On-site
Employment Type
Full-time
Categories
AI Agent Engineer
About the job
This role is for a Lead Software Engineer focused on building an AI Agent Platform. You will lead the engineering efforts for a high-visibility platform that includes an SDK, orchestration frameworks, and reusable components to power agentic AI experiences across the organization. Your responsibilities will include owning the foundational tooling for composing agents, grounding them in enterprise data, providing durable memory, and improving accuracy on complex capabilities like NL2SQL and RAG. You will also be responsible for raising delivery standards, turning complex problems into reliable production systems, and setting the quality bar for infrastructure used by many teams and users. The team values strong engineering, thoughtful collaboration, and continuous learning.
Responsibilities
- Lead end-to-end delivery of the AI Agent Platform SDK and frameworks, from problem framing and technical design through production deployment, scaling, and monitoring.
- Own core platform services and reusable building blocks: agent orchestration, tool/function calling, retrieval-augmented generation pipelines, and model-serving integration.
- Build and operate first-class agent modalities on the platform, including voice agents and document-extraction agents.
- Drive systematic improvement of core agent capabilities like NL2SQL, RAG, voice, document extraction, and tool use by building evaluation datasets, benchmarks, and quality metrics.
- Establish engineering standards through system design, code review, and mentorship to improve reliability, maintainability, latency, and developer experience.
- Build and operationalize evaluation, testing, and observability capabilities to continuously improve solution quality.
- Implement robust safety and governance patterns, including guardrails, content filtering, prompt-injection defenses, and access controls.
- Partner with product managers and stakeholders to shape the roadmap, define success metrics, and prioritize work.
- Drive cross-functional alignment across engineering, data, security, and risk partners.
- Contribute to technical documentation, SDK references, reference implementations, and enablement content.
Requirements
- Formal training or certification on applied artificial intelligence and machine learning concepts and 5+ years applied experience.
- Advanced proficiency in Python with strong software engineering fundamentals (testing, design patterns, version control, code review).
- Hands-on experience building, evaluating, and deploying ML or LLM-enabled systems into production environments (ideally libraries, SDKs, or platform services).
- Practical experience with prompt engineering and retrieval-augmented generation, including building evaluation methods and quality measurement.
- Experience designing and operating reliable services, including incident response readiness, performance tuning, and operational stability for data-intensive systems.
- Demonstrated ability to lead technical decisions and deliver outcomes through ambiguity, balancing speed, risk, and long-term maintainability.
- Strong communication skills with the ability to explain technical trade-offs to both technical and non-technical stakeholders.
- Experience with agent orchestration frameworks (e.g., LangGraph, LlamaIndex, Google ADK, or custom orchestration).
- Experience building voice agents (speech recognition, text-to-speech, real-time streaming audio, conversational turn-taking).
- Experience building document-extraction / intelligent document-processing agents (OCR, layout parsing, structured field and table extraction).
- Experience implementing agentic memory systems (short- and long-term memory, persistence, context retrieval).
- Familiarity with vector databases, embedding pipelines, or graph-based memory approaches.
- Experience with continuous integration.