Sr. Software Engineer

Overview

Build, deploy and operate production AI agents and retrieval solutions on AWS (Amazon Bedrock and AgentCore) and Oracle (AI Data Platform and AI Database) for JLL's enterprise data.

What you'll do

• Design, deploy, and operate AI agents on Amazon Bedrock AgentCore, using Runtime, Gateway (APIs and Lambda functions as tools, including MCP), Memory, Identity, and Observability.
• Build RAG and agent solutions with Amazon Bedrock, implementing foundation-model integrations, Knowledge Bases, Agents, and Guardrails, and manage prompts and model selection for cost, latency, and quality.
• Build on Oracle AI Data Platform (AIDP): develop in the AIDP Workbench using Master Catalog assets, Spark notebooks, and workflows; apply medallion (bronze, silver, gold) data patterns; and build knowledge bases and agent flows on governed enterprise data.
• Build on Oracle AI Database: implement AI Vector Search, hybrid (vector + relational) retrieval, Select AI natural-language-to-SQL interfaces, and in-database ML, with embedding pipelines and PL/SQL where needed.
• Integrate Oracle data services with AWS-hosted AI applications through secure, well-defined APIs and swappable service interfaces.
• Engineer retrieval quality: own chunking and embedding strategy, metadata filtering, reranking, and retrieval tuning, and measure results with repeatable evaluations.
• Secure and govern AI: apply IAM least privilege, database-enforced access control (such as role-based access and data redaction), PII handling, and guardrails that screen for prompt injection and unsafe output; keep audit trails of agent actions and outcomes.
• Test and evaluate: build evaluation and regression suites for LLM output, track hallucination and accuracy, and verify behavior end to end against real services, not only with unit tests.
• Deploy and operate Python services (for example FastAPI) to production on AWS using Lambda, API Gateway, S3, ECS/EKS, Step Functions, and Secrets Manager; monitor latency, quality, and cost with CloudWatch and AgentCore Observability.
• Collaborate and document: take part in sprint ceremonies and code and design reviews, and maintain documentation, runbooks, and knowledge-transfer material.

What you'll need

• 4-8 years of software, data, or AI/ML engineering experience.
• Hands-on delivery of generative AI or LLM-based applications in production or near-production.
• Proven experience building RAG pipelines and agentic workflows, including tool calling and multi-step orchestration.
• Experience with Amazon Bedrock and AWS-native deployment.
• Experience with Oracle Database or Oracle AI Database.
• Strong debugging and root-cause analysis skills, including tracing non-deterministic model behavior.
• Clear written and verbal communication, and the ability to work independently in a distributed team.
• Programming in Python (primary), REST APIs, and SQL; FastAPI or similar async frameworks.
• AWS cloud: Lambda, API Gateway, S3, ECS/EKS, Step Functions, IAM, VPC, CloudWatch, and Secrets Manager.
• Security and governance: IAM least privilege, data privacy controls, PII handling, and responsible-AI guardrails.
• Delivery: Agile/Scrum, Git and CI/CD, Docker, Jira and Confluence, and clear documentation.

Nice to have

• Bachelor's degree or higher in Computer Science, AI, or a related field.
• Hands-on AgentCore experience.
• Experience with the Oracle AI Data Platform.
• Experience with AgentCore Identity and Gateway for secure, tool-enabled agents.
• Experience with Oracle Cloud Infrastructure (OCI), Apache Spark, or medallion-architecture data pipelines.
• Infrastructure as code (for example Terraform) and Kubernetes experience.
• Experience with agent frameworks such as Strands Agents, or with migrating Bedrock Agents to a code-first framework.
• Experience with LLM gateways, token and cost tracking, and usage analytics.
• Relevant certifications, such as AWS or Oracle Cloud Infrastructure Generative AI certifications.

Details

• Location: Remote, Bengaluru, KA.
• Scheduled weekly hours: 40.