Amazon Connect application development and configuration: Build, update, and maintain Amazon Connect contact flows, routing profiles, and queues to support both voice and chat channels. Integrate Lex V2 bots into Amazon Connect flows for call deflection, self-service transactions, and escalation. Enable and configure Contact Lens for real-time and post-contact analytics, transcription, summarization, sentiment analysis, and redaction. Configure Amazon Q in Connect domains, knowledge sources, guided workflows, and step-by-step agent assist experiences. Implement S3-based call and chat transcript storage with encryption, lifecycle policies, and retention compliance. AI orchestration and backend services: Write AWS Lambda (Python) functions to orchestrate Bedrock LLM calls, embeddings workflows, and model invocation logging. Implement OpenSearch indexing, vector/keyword queries, and knowledge synchronization triggers. Build retrieval-augmented generation (RAG) pipelines to enhance Amazon Connect agent assist and self-service knowledge. Apply structured logging, unit/integration tests, error handling, and performance/cost safeguards. Data safety and compliance: Implement AI guardrails, prompt templates, and output evaluation for safety and accuracy. Enforce PII minimization and redaction policies in Amazon Connect conversation logs. Participate in threat modeling and support remediation of findings for contact center integrations. Web and chat integration: Support CloudFront + WAF configurations for secure web chat entry points. Build APIs and event hooks to pass conversation context between web chat, Amazon Connect, and AI services. DevSecOps and operations: Contribute to Git-based CI/CD pipelines, code reviews, and documentation. Maintain runbooks, architecture diagrams, and SOPs for contact flows, bot integrations, and AI workflows. Create and monitor CloudWatch dashboards/alarms for call deflection rate, average handle time (AHT), contact resolution, and AI usage metrics.
