A complete application in a minute
Tailored resume and cover letter, ready to send.
Globant is seeking an AI/ML Engineer with 3–5 years of experience to design, build, andProductionize pricing agents within a federated agent portfolio. You will own end-to-end workflows from prototype to production, including Graph RAG retrieval, agent memory design, and cross-system data integration.
The role requires strong Python, ML, and MLOps skills, with experience in Databricks or SageMaker, Kubernetes, and real-time data processing.
Salary: $100,000-$120,000
Design and ship production pricing agents on NEO, owning them from prototype through production, as part of the firm's federated portfolio of production agents. Build retrieval that holds up in production: Graph RAG combining knowledge-graph traversal over products, rate cards, clients, and pricing policy with vector search, plus chunking, ranking, and grounding strategies that keep price explanations accurate and auditable. Design agent memory: episodic and semantic memory organized as memory nodes, with recall, summarization, and decay policies tuned per use case, so an agent remembers the history of a negotiation without leaking it across clients. Own organizational context management - assembling entitlement-, lineage-, and tenant-aware context so each agent reasons over only the pricing data it is allowed to see. Build and productionize ML model training pipelines for price optimization, elasticity, win-probability, and deal-scoring models using MLOps best practices, including feature engineering over billing, transaction, and deal history data on Databricks. Build batch and online inference for those models, meeting the latency a seller conversation requires while keeping every scored recommendation reproducible. Own model validation in a revenue-sensitive setting: back-testing against realized prices, challenger models, uplift and elasticity measurement, and monitoring for drift, margin erosion, and revenue leakage in production. Compose multi-agent workflows using A2A, and integrate tools and data through MCP servers (pricing engine, rate card store, CRM, billing and revenue data, Databricks, Snowflake, Kubernetes, Confluence, Bitbucket, Splunk). Build and run evals: task-level and end-to-end agent evaluations, regression suites, LLM-as-judge, numerical-correctness and explanation-faithfulness checks, and quality and safety gating before release. Produce the model documentation, test evidence, and controls the firm's model risk review requires for a model that influences client pricing, and keep that evidence current as models are retrained. Deploy and operate solutions on public cloud (AWS and/or Azure) with strong SDLC, security, resiliency, and observability practices. Partner with pricing, product, finance, and business partners across CIB and Payments to turn use cases into shipped, supported agents.