Job description

Founded in 2004, NetBrain is the leader in no-code network automation. Its ground-breaking Next-Gen platform provides IT operations teams with the ability to scale their hybrid multi-cloud connected networks by automating the processes associated with Diagnostic Troubleshooting, Outage Prevention and Protected Change Management. Today, over 2,500 of the world’s largest enterprises and managed services providers leverage NetBrain’s platform. What We Need We’re looking for a Senior AI Engineer to design and build production-grade agent and RAG systems that power intelligent, reliable automation across our platform. This role combines hands-on engineering with system-level thinking—owning everything from architecture and evaluation to scalability, observability, and reliability in production. The ideal candidate thrives in ambiguity, moves quickly from prototype to production, and brings a strong focus on quality, safety, and real-world impact. What You'll Do Agent Platform Architecture Design and implement core capabilities for an enterprise-grade Agent platform, including orchestration patterns such as ReAct, Plan-and- Execute, and Supervisor, as well as tool execution, context and memory management, and safety guardrails. Design enterprise-grade Agent execution and governance mechanisms, including Human-in-the-Loop approval workflows, multi-tenant permission isolation, policy enforcement, and secure execution controls. Build reusable Agent Skills, standardized tool interfaces, and a scalable tool ecosystem deeply integrated with NetBrain platform capabilities and business workflows. LLM and Model Optimization Design and implement LLM post-training strategies, including domain-specific Supervised Fine-Tuning (SFT), DPO/RLHF-based preference alignment, and parameter-efficient fine-tuning techniques such as LoRA, to continuously improve model performance in the network operations domain. Build an Agent self-learning feedback loop that converts production execution traces, user feedback, and evaluation results into high- quality datasets for continuously improving prompts, skills, models, and retrieval strategies. Analyze and optimize LLM behavior across areas such as instruction following, tool calling, structured output generation, contextual understanding, reasoning stability, and hallucination mitigation. Evaluation, Reliability, and Observability Build production-grade LLM and Agent evaluation frameworks and automated regression pipelines, including benchmark datasets, deterministic checks, LLM-as-a-Judge, tool-call validation, retrieval-quality evaluation, and end-to-end task success metrics. Establish release quality gates and hallucination-detection mechanisms for AI features to prevent significant accuracy, reliability, and performance regressions from reaching production. Build comprehensive AI system observability capabilities, including distributed tracing, structured logging, metrics, dashboards, and alerting. Rapidly diagnose and resolve production AI failures, including hallucinations, incorrect tool selection, invalid tool parameters, Agent loops, retrieval-quality degradation, structured-output failures, latency regressions, and unexpected model behavior changes. Production Engineering and Technical Execution Design and implement highly reliable backend services for production AI and Agent workloads, including asynchronous and concurrent processing, retries, timeouts, caching, rate limiting, and fault isolation. Continuously optimize latency, throughput, token consumption, and infrastructure cost to meet platform SLA requirements and support large-scale production workloads. Independently diagnose and resolve complex AI system issues spanning prompts, models, RAG, tools, Agent workflows, backend services, and infrastructure. Lead technical design for critical modules and system-level capabilities, ensuring solutions align with platform architecture, security requirements, engineering standards, and product requirements. Drive technical improvements based on production data, evaluation results, and benchmarks, and collaborate closely with Engineering, Product, QA, and other teams to deliver solutions into production. Applied Research and Technical Strategy Prototype, benchmark, and productionize emerging technologies such as GraphRAG, Knowledge Graphs, MCP, LLM Post-Training, and Agent Self-Learning to improve grounding, multi-hop reasoning, and domain expertise. Continuously evaluate Agent frameworks and supporting infrastructure, including LangChain, LangGraph, AutoGen, and LlamaIndex, and provide technical recommendations for platform architecture evolution and product technology strategy. Stay current with developments in LLM and Agent technologies and rapidly translate promising technologies into measurable, testable, and production-ready engineering capabilities. What You Bring Bachelor's degree or higher in Computer Science, Artificial Intelligence, Electrical Engineering, or a related technical field. Master's or Ph.D. preferred; equivalent practical experience will also be considered. 3+ years of experience in software engineering, machine learning, or applied AI, including 2+ years building, deploying, and operating production- grade LLM or Agent applications. Must have delivered at least one LLM-powered feature end-to-end and owned its ongoing operation and improvement after production launch. Deep understanding of Agent architectures and LLM behavioral characteristics, including instruction following, tool-calling behavior, and context sensitivity, with hands-on experience building multi-step workflows involving reasoning, tool execution, state management, structured outputs, validation, and error recovery. Proven ability to diagnose and resolve production LLM/Agent failures, including hallucinations, incorrect tool calls, retrieval-quality degradation, Agent loops, structured-output failures, latency regressions, and regressions introduced by prompt or model changes. Strong Python and distributed backend engineering skills, including API and service development, asynchronous and concurrent programming, retries, timeouts, caching, rate limiting, testing, logging, and cross-service performance debugging. Hands-on experience designing evaluation systems for LLM applications, including dataset construction, metric definition, regression testing, and release quality gates. Strong understanding of security risks associated with LLM and Agent applications, including prompt injection, data leakage, unsafe tool execution, permission boundaries, and uncontrolled Agent autonomy. Ability to independently design, implement, debug, deploy, and operate complex production systems. Preferred Qualifications Strong experience with RAG and advanced retrieval systems, including embeddings, vector and hybrid search, reranking, chunking strategies, grounding and citation mechanisms, as well as multi-hop retrieval, Knowledge Graphs, or GraphRAG. Familiarity with Agent frameworks such as LangGraph, LangChain, AutoGen, and LlamaIndex, as well as standardized protocols such as MCP (Model Context Protocol) for connecting Agents with external tools and systems. A strong understanding of the underlying architecture is more important than expertise in any specific framework.Experience designing Agent runtime mechanisms, including Human-in-the-Loop workflows such as risk classification, approval, interruption, pause /resume, as well as context management for long-running Agent workflows, including state persistence, history compression, and memory systems. Experience with LangSmith or similar LLM observability and evaluation platforms. Experience with LLM fine-tuning, including LoRA or other parameter-efficient fine-tuning techniques. Experience applying LLM technologies to networking, infrastructure, cybersecurity, observability, or other complex technical domains. Fluent in both English and Chinese, with strong cross-regional communication and collaboration skills. What We Offer Our comprehensive compensation package is vital in how we recognize our people for the impact they make on us reaching our goals as a company. For this role, the estimated base is $150,000 - $180,000 + Bonus. The actual salary may vary based on a range of factors, including market and individual qualifications objectively assessed during the interview process. The range listed above is a guideline and may be modified. People Experience offers a comprehensive benefits package in addition to cash compensation that includes but is not limited to 401k and medical/dental coverage. Speak with your Recruiter for more details on our Total Rewards philosophy. #LI-BW1 NetBrain invites all interested and qualified candidates to apply for employment opportunities. Qualified applicants will receive consideration for employment without regard to race, color, religion, sex, national origin, disability, protected veteran status, or other characteristics protected by law. If you have a disability that prevents or limits your ability to use or access the site, or if you require any other accommodation in the application process due to a disability, you may request a reasonable accommodation. To make a request, please contact our People Team at: people@netbraintech.com and we will be happy to assist you. In compliance with applicable laws, NetBrain conducts holistic, individual background reviews in support of all hiring decisions. It is unlawful in Massachusetts to require or administer a lie detector test as a condition of employment or continued employment. An employer who violates this law shall be subject to criminal penalties and civil liability.

Requirements

Must-haves

Agent Platform Architecture
Design and implement core capabilities for an enterprise-grade Agent platform, including orchestration patterns such as ReAct, Plan-and-
Execute, and Supervisor, as well as tool execution, context and memory management, and safety guardrails.
Design enterprise-grade Agent execution and governance mechanisms, including Human-in-the-Loop approval workflows, multi-tenant
permission isolation, policy enforcement, and secure execution controls.
Build reusable Agent Skills, standardized tool interfaces, and a scalable tool ecosystem deeply integrated with NetBrain platform
capabilities and business workflows.
LLM and Model Optimization
Design and implement LLM post-training strategies, including domain-specific Supervised Fine-Tuning (SFT), DPO/RLHF-based
preference alignment, and parameter-efficient fine-tuning techniques such as LoRA, to continuously improve model performance in the
network operations domain.
Build an Agent self-learning feedback loop that converts production execution traces, user feedback, and evaluation results into high-
quality datasets for continuously improving prompts, skills, models, and retrieval strategies.

Nice-to-haves

preferred; equivalent practical experience will also be considered.
Strong experience with RAG and advanced retrieval systems, including embeddings, vector and hybrid search, reranking, chunking strategies,
grounding and citation mechanisms, as well as multi-hop retrieval, Knowledge Graphs, or GraphRAG.
Familiarity with Agent frameworks such as LangGraph, LangChain, AutoGen, and LlamaIndex, as well as standardized protocols such as MCP
(Model Context Protocol) for connecting Agents with external tools and systems. A strong understanding of the underlying architecture is more
important than expertise in any specific framework.Experience designing Agent runtime mechanisms, including Human-in-the-Loop workflows such as risk classification, approval, interruption, pause
/resume, as well as context management for long-running Agent workflows, including state persistence, history compression, and memory
Experience with LangSmith or similar LLM observability and evaluation platforms.
Experience with LLM fine-tuning, including LoRA or other parameter-efficient fine-tuning techniques.
Experience applying LLM technologies to networking, infrastructure, cybersecurity, observability, or other complex technical domains.
Fluent in both English and Chinese, with strong cross-regional communication and collaboration skills.

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

aireactllmmachine learningpythonapi

Platform: greenhouse
Effort: quick

Benefits

Benefits detected in the listing.

Dental

401k