LogoFresh Jobs Ahead
Equifax
Equifax
H1B vizos rėmėjasFinansaiFintechKibernetinis saugumas
„Equifax“ yra pasaulinė finansų, finansinių technologijų ir kibernetinio saugumo bendrovė, siekianti duomenis paversti įžvalgomis, padedančiomis priimti svarbius sprendimus. Jos paslaugos apima kredito ataskaitas, rizikos modeliavimą, tapatybės apsaugą, analitiką, debesijos technologijas ir dirbtinį intelektą, skirtus finansų įstaigoms, darbdaviams ir valdžios institucijoms. Atlantoje įsikūrusi „Equifax“ turi beveik 15 000 darbuotojų 24 šalyse ir derina nuosavybės teise valdomus duomenis su technologijomis, kad padėtų klientams priimti informacija pagrįstus sprendimus.
LinkedIn
Darbai įmonėje

Software Engineer II, Agentic AI

Build production LLM agents, retrieval-augmented generation pipelines, and enterprise AI workflows at Equifax. Develop resilient systems using Python, orchestration frameworks, APIs, and vector databases.

INIndijaHibridinisVisas etatasVidutinis lygisVyresnysis specialistasPrivaloma Kalba: anglųFull-stack inžinieriusPrograminės įrangos inžinierius
DebesijaDockerGoogle Cloud PlatformgRPCPythonReact

Aprašymas

  • Build stateful, multi-step agent workflows with frameworks such as LangGraph, AutoGen, or LlamaIndex.
  • Improve prompts with techniques such as Chain-of-Thought (CoT) and ReAct.
  • Create robust API wrappers and strict JSON schemas for LLM tool and function calls.
  • Connect agents securely to internal databases, CRMs, and enterprise systems.
  • Develop context retention using short-term memory buffers and long-term vector or relational storage.
  • Build and maintain high-throughput data ingestion pipelines.
  • Improve search relevance through hybrid search, embedding tuning, and reranking.
  • Add error handling and retries for rate limits, API failures, and hallucinations.
  • Enable agents to correct their own outputs or escalate clearly to human operators.
  • Write deterministic software tests and evaluations using LLM-as-a-judge methods.
  • Use tracing tools such as LangSmith and Phoenix to inspect reasoning paths, prevent infinite loops, and reduce latency.
  • Turn AI prototypes into reliable production systems.
  • Write clean, efficient code; refine LLM prompts; integrate APIs; and manage stateful agent workflows.

Reikalavimai

  • Bring 3–6 years of core software engineering experience.
  • Have at least one year of hands-on experience building and deploying production LLM applications, RAG pipelines, or agentic workflows.
  • Use Python proficiently, including FastAPI, Pydantic, and Asyncio.
  • Have hands-on experience with LLM orchestration frameworks such as LangGraph, LangChain, AutoGen, or LlamaIndex.
  • Have experience writing JSON schemas for function calling.
  • Have experience building REST or gRPC API wrappers.
  • Have experience working with vector databases.
  • Have experience with observability platforms for tracing agent execution, monitoring latency, and evaluating prompts.
  • Bring a solid foundation in CI/CD, Git workflows, unit and integration testing, Docker, and clean, scalable software development.
  • Have hands-on experience with GCP cloud-native AI services, including Vertex AI, ADK, and Model Armour.
  • Have exposure to semantic guardrail frameworks such as NeMo Guardrails or Guardrails AI.
  • Be familiar with fine-tuning, embedding customization, or deploying open-source models locally.

Privalumai

  • Comprehensive compensation and healthcare packages.
  • Paid time off.
  • Career growth opportunities through an online learning platform with guided career tracks.
  • Hybrid work setting.

Susiję darbai