AI Software Engineer — Agentic AI & Copilot Systems

Globant; New Jersey; On-site; USD 85,000 - 120,000; Full time

Job summary

Globant is hiring an AI Software Engineer to build secure, production-ready AI agents and copilot capabilities. You will join a team delivering intelligent workflows across research, summarization, and follow-up activities, with a focus on security, reliability, and responsible AI.

You will design agent components, develop MCP servers, and implement retrieval pipelines, data processing on Databricks, and MLOps workflows from development to UAT.

Qualifications

• Formal training or certification in software engineering concepts and at least 3 years of applied experience.
• Hands-on experience in system design, application development, testing, and operational stability.
• Proficiency in Python and at least one additional language, such as Java, TypeScript, or SQL.
• Experience building and consuming APIs and event-driven services in a cloud environment.
• Experience with data processing frameworks such as Apache Spark and large structured datasets.
• Working knowledge of LLM application development, including prompting, tool calling, retrieval-augmented generation (RAG), and evaluation.
• Experience using approved AI-assisted software development tools, with sound judgment when validating their outputs.
• Understanding of responsible AI practices, data sensitivity, and secure handling of inputs and outputs.
• Solid understanding of agile delivery practices, CI/CD, application resiliency, and security.
• Practical cloud-native experience, including containers and Kubernetes; AWS experience is preferred.

Responsibilities

• Design, develop, and troubleshoot agent components, including orchestration logic, tools, and task queues supporting research, summarization, and follow-up workflows.
• Build and maintain Model Context Protocol (MCP) servers that provide secure access to CRM systems, client and product reference data, document stores, calendars, and email, enforcing user entitlements on every call.
• Write secure, high-quality production code in Python, and review, debug, and improve code written by others.
• Develop retrieval pipelines for product, pricing, legal, and client documents, including ingestion, chunking, embedding, indexing, and citation metadata to ensure answers are traceable to their sources.
• Build data pipelines and feature engineering jobs on Databricks to prepare CRM, transaction, and engagement data for model training and copilot retrieval.
• Contribute to MLOps workflows for propensity, next-best-action, and opportunity-scoring models, including training jobs, automated testing, and promotion of model artifacts from development through user acceptance testing (UAT) to production.
• Develop the copilot’s user-facing experience and streaming interactions using AG-UI, including human-in-the-loop review before communications are sent to clients or changes are written to CRM.
• Create agent evaluations, including offline test sets, groundedness and citation checks, CI/CD regression suites, and LLM-as-judge scoring.
• Instrument services with OpenTelemetry tracing to support observability and troubleshooting.
• Use enterprise-authorized AI-assisted development tools while validating generated outputs for correctness, performance, and security.
• Identify opportunities to eliminate recurring issues or automate remediation to improve the operational stability of agents and model services.
• Contribute to a team culture rooted in diversity, opportunity, inclusion, and respect.

Skills

Python

Java

TypeScript

SQL

APIs

Cloud

Kubernetes

AWS

CI/CD

Security best practices

Tools

Databricks

OpenTelemetry

Terraform

AG-UI

Job description

Globant is hiring an AI Software Engineer to build secure, production-ready AI agents and copilot capabilities. You will join a team delivering intelligent workflows across research, summarization, and follow-up activities, with a focus on security, reliability, and responsible AI.

You will design agent components, develop MCP servers, and implement retrieval pipelines, data processing on Databricks, and MLOps workflows from development to UAT.