Senior Software Engineer

Overview

Senior Software Engineer on the AI Engineering team in Bengaluru. Embeds with client teams to understand manual business workflows and builds AI-agent-powered automation to replace them, helping build a new agent harness platform and delivering production workflows on it.

What you'll do

• Embed with client teams and understand manual business workflows to replace them with AI-agent-powered automation.
• Help build a new agent harness platform from the ground up on Claude (Claude Code / Cowork), AWS Step Functions (ASL) and Amazon Bedrock AgentCore, with an AG-UI front end.
• Own the solution end-to-end: discovery, build, deploy, evaluate and iterate, working directly with enterprise customers with minimal direction.
• Design agent harnesses covering the control loop, tool and skill surface, context and state management, and human-in-the-loop checkpoints.
• Author Amazon States Language state machines to orchestrate agent steps, branching, retries, parallelism and human-approval gates.
• Design workflows where policy gates and rules decide routing and the LLM only proposes, with explicit exception, timeout and loop-guard paths.
• Host, operate and scale agent workloads on AgentCore Runtime and expose tools via AgentCore Gateway.
• Trace, measure and log agent runs with a per-step audit trail to debug and monitor production workflows.
• Design and run evaluations to measure agent quality before and after each release.
• Configure Amazon Bedrock Guardrails and policy controls for safety, prompt-injection defence, PII and hallucination mitigation.
• Design, deploy and scale AI-powered production workflows using Claude Code and/or Cowork.
• Build and consume MCP servers and integrate tools so agents reach data only through MCP/API.
• Build agent-facing user interfaces on the AG-UI protocol for streaming responses, tool-call visibility, shared state and human approval steps.
• Decompose manual business processes into agent workflows, including document processing and classification.
• Extract text, tables and layout from scanned documents using OCR and feed the output into agent workflows.
• Integrate enterprise systems such as SharePoint, email and external systems via APIs and MCP.
• Provision and promote AgentCore, Step Functions, Lambda and supporting AWS resources using Terraform.

What you'll need

• Excellent understanding of harness engineering for agent control loops, tools, skills, context and state management, and human-in-the-loop checkpoints around an LLM.
• Working knowledge of AWS Step Functions with ASL, including task-token waits and invoking AgentCore runtimes and Lambda from workflows.
• Hands-on experience with Amazon Bedrock AgentCore, including Runtime lifecycle, scaling and invocation patterns.
• Hands-on tracing, metrics and logging of agent runs using AgentCore, CloudWatch or Datadog.
• Hands-on design and running of evaluations such as AgentCore Evaluations or golden-dataset regression tests.
• Hands-on configuration of Amazon Bedrock Guardrails and policy controls.
• Experience designing, deploying and scaling AI-powered production workflows using Claude Code and/or Cowork.
• LLM API fundamentals including tool use, function calling, streaming and context management.
• Experience designing and publishing reusable skills consumable by agent workflows.
• Experience building, hosting and consuming MCP servers.
• Experience with multi-agent coordination, including task delegation and result aggregation (A2A).
• Experience with RAG pipelines, including vector databases, chunking, embeddings and retrieval.
• Experience with prompt engineering, including system prompts, evaluation and iteration.
• Experience building AG-UI interfaces.
• Python 3.10+ with async/await and type hints.
• FastAPI or an equivalent Python web framework.
• Basic SQL to query, filter and join, with awareness of vector databases such as pgvector, Pinecone or OpenSearch.
• Working knowledge of document or key-value stores such as Amazon DynamoDB, including data modelling and access patterns.
• REST API design and JWT / OAuth2.
• Document processing skills covering PDF, Excel and unstructured data extraction and classification.
• OCR experience, for example with AWS Textract.
• Enterprise system integration via APIs and MCP.
• AWS services including Lambda, ECS, RDS, S3 and IAM.
• Terraform / Infrastructure as Code.
• Docker and containerisation.
• GitHub Actions CI/CD.
• Proven software engineering experience with AI engineering shipped to production.
• At least one production project involving LLM agents, skill building or RAG pipelines, ideally on AWS Bedrock AgentCore and/or Step Functions.
• Direct experience working with enterprise customers to design, deploy and scale AI-powered production workflows using Claude Code / Cowork.
• Track record of owning and delivering solutions end-to-end with minimal guidance.
• Experience working directly with non-technical stakeholders, gathering requirements and presenting solutions.

Nice to have

• Kubernetes.
• TypeScript.
• Fine-tuning and model evaluation.
• CRE domain knowledge.
• Consulting or client-facing background.

Details

• Location: On-site, Bengaluru, KA.
• Scheduled weekly hours: 40.