C2C requirements
Location: Naperville, IL (Onsite position) Locals only
Duration: 6+ Months
Responsibilities
Build RAG-based chatbots over S3-hosted knowledge: ingestion, parsing, chunking, metadata enrichment, embeddings, indexing, retrieval, reranking, and citation-grounded responses.
Implement Amazon Bedrock model integration and Bedrock Guardrails (PII handling, content safety, prompt-injection defenses, topic restrictions, safe completion).
Develop agentic capabilities using Bedrock Agents/Agent Core and Lambda tools: allowlisted tool calling, schema validation, approvals, audit logging, and safe execution patterns.
Engineer AWS-native pipelines for KB updates and automation using S3 eventing, orchestration, and resilient processing (retries, DLQs, idempotency).
Design and maintain LLM evaluations and quality gates: golden datasets, regression testing in CI/CD, grounded Ness/citation checks, retrieval quality metrics, latency/cost reporting.
Implement production-grade observability and operations: CloudWatch dashboards/alarms, structured logging/tracing, incident readiness, and cost governance.
Must-Have Qualifications (All Required)
Strong Python engineering background with production experience in cloud-native services.
Extensive hands-on experience with AWS services:
Amazon Bedrock (FMs + embeddings) and Bedrock Guardrails
AWS Lambda, CloudWatch, DynamoDB (schema and access patterns)
S3 (bucket architecture, automation, event-driven ingestion)
Amazon Kendra (enterprise search / hybrid retrieval patterns)
AWS Agent Core / Bedrock Agents (agent orchestration and tool use)
Proven expertise in:
LLMs (prompting patterns, context management, hallucination mitigation)
RAG (chunking, metadata, hybrid retrieval, reranking, citations)
Data engineering on AWS (pipelines, quality checks, lineage, versioning)
Database schema awareness and designing data access patterns for scalable retrieval
Production delivery skills:
API Gateway, Step Functions, Event Bridge, SQS/SNS
OpenSearch / vector database patterns (as applicable)
IAM least privilege, KMS, Secrets Manager/SSM, CloudTrail
CI/CD and IaC (Terraform/CloudFormation/CDK)
Reliability patterns (rate limiting, caching, retries, circuit breakers, DR awareness)
AI certification required (AWS ML Specialty or equivalent recognized AI/ML certification) and strong depth across AWS services.
To apply for this job email your details to narayana@absli.com