Gopi Kiran Manusani

Full-stack GenAI engineer with 16+ years of enterprise IT experience, specializing in deploying and scaling agentic AI tooling across enterprise SDLC workflows. Hands-on daily with Claude Code and Cursor for agent-driven development, and architect of multi-agent platforms (LangGraph, AutoGen, CrewAI), enterprise RAG pipelines on Amazon Bedrock, and tool-calling/ReAct agent frameworks that integrate GenAI directly into CI/CD and production operations. Proven track record partnering with line-of-business teams to drive tool onboarding, training, and adoption — cutting MTTR ~45% and enabling 60%+ autonomous ticket resolution.

Full-stack builder (Java, Python, Node.js, React.js) across distributed, event-driven microservices on AWS and hybrid AWS–GCP infrastructure, with AWS Certified Solutions Architect credentials and a background in regulated, governance-heavy enterprise environments.

Generative AI Engineer | Agentic AI Platform Engineer | Full-Stack & GenAI Enablement Lead - COMCAST - Atlanta, GA (Remote)

(2024-02)

• Drove hands-on adoption of AI-native developer tooling — Cursor, Claude Code (CLI), OpenAI Codex, and GitHub Copilot/Continue.dev — for agent code generation, test scaffolding, and review automation, partnering directly with NOC and platform teams to fold GenAI into daily delivery workflows.
• Led tool onboarding, training, and best-practice enablement for engineers adopting agentic coding assistants and LLM-powered workflows, establishing prompt-engineering standards that improved consistency and quality of AI-assisted output across teams.
• Established LangSmith tracing of token usage, chain paths, and agent decisions across all pipelines to monitor tool usage, efficiency, and performance outcomes — feeding continuous optimization of agent behavior and cost.
• Authored architecture blueprints and mentored junior engineers on LLM patterns, agent design, and AWS best practices, translating business requirements from line-of-business stakeholders into production-ready GenAI solutions.
• Architected a multi-agent AI platform (LangGraph + AutoGen) automating L1/L2 NOC incident triage — agents retrieve runbooks, query CloudWatch, correlate events, and draft remediations via tool-chaining and decision-driven execution, cutting MTTR ~45%.
• Built enterprise RAG pipelines on Amazon Bedrock (Claude 3) with LlamaIndex and OpenSearch vector search, powering natural-language Q&A over 5,000+ runbooks and compliance documents.
• Implemented ReAct agents with Lambda tool calling — autonomous, decision-driven workflows for subscriber provisioning, config-change validation, and rollback — removing manual NOC intervention for routine work.
• Shipped an AI helpdesk chatbot (Lex + Bedrock LLM + DynamoDB sessions) autonomously resolving 60%+ of Tier-1 tickets; built a GenAI fault-prediction system with fine-tuned SageMaker models and Comprehend.
• Designed a document intelligence system (Textract + Bedrock + pgvector on RDS) for semantic search across technical specs, contract SLAs, and audit logs.
• Enforced responsible AI with Bedrock Guardrails and prompt standards to keep agentic workflows compliant in a governance-heavy enterprise environment.
• Built Node.js REST APIs and React.js dashboards exposing AI insights and network metrics to NOC and LOB teams, integrating agentic backend services with full-stack front-end delivery.
• Owned end-to-end CI/CD across CodePipeline/CodeBuild/CodeDeploy, GitHub Actions, and GitLab CI — automated build, test, security scan, and deploy stages delivering zero-downtime blue/green and canary releases for AI applications and network services.
• Extended pipelines to treat prompts, model configs, and agent definitions as code — versioned in Git, gated by automated LLM evaluation suites, and released through the same review/deploy/rollback lifecycle as application code, integrating GenAI directly into the enterprise SDLC.
• Embedded DevSecOps controls shift-left in every pipeline: SAST/DAST scanning, container image vulnerability scanning before ECR promotion, IaC policy checks, and Vault-injected secrets.
• Implemented GitOps-style delivery for Amazon EKS — declarative manifests versioned in Git as the single source of truth, peer-reviewed pull-request promotion across Dev/QA/Prod, and automated rollback on CloudWatch alarm breach.
• Centralized observability with CloudWatch, CloudTrail, and the ELK Stack — dashboards and alerting for deployment health, AI performance (latency, token throughput, error rates), and availability SLAs; tracked release frequency, lead time, and change-failure rate.
• Provisioned scalable AWS infrastructure as code (VPC, subnets, EC2, ELB/ALB, Auto Scaling, Security Groups) with Terraform and CloudFormation — reusable, versioned modules supporting ISP network services and AI/ML inference workloads.
• Operated Amazon EKS microservices — subscriber provisioning, usage metering, distributed AI inference serving — with horizontal pod autoscaling and dedicated GPU node pools for ML workloads.
• Built event-driven serverless automation (Lambda, Step Functions, EventBridge) for subscriber lifecycle management, agentic task orchestration, and alarm-triggered incident response across distributed systems.
• Delivered cross-cloud AWS–GCP integration for hybrid analytics and model-training pipelines supporting network traffic analysis and capacity planning.
• Hardened the platform with least-privilege IAM, Security Group governance, HashiCorp Vault, and TLS/SSL, extended to protect API keys and model-access credentials for all GenAI services.

Java/Spring Boot Solutions Architect | IBM ODM & BPM Tech Lead | Business Analyst | AWS Cloud SME - ELI LILLY (VIA TCS) - Indianapolis, IN

(2015-10 - 2024-01)

• Architected and engineered Java 8/11, Spring Boot, and Spring Cloud microservices exposing IBM ODM decision services and business rule APIs via JAX-RS/REST to BPM workflows, SharePoint, and downstream enterprise applications.
• Served as Solutions Architect for the decisioning platform — defined service boundaries, domain models, and integration contracts across a Java/Spring Boot microservices architecture, establishing coding standards, design patterns, and architecture governance for onshore/offshore delivery teams.
• Engineered Java XOM domain models and reusable Spring Boot components, applying Spring Security, Spring Data JPA, and RESTful API design to standardize decision-service integration across regulatory, clinical, and HR platforms.
• Led ODM migration v8.5 → v8.9.2 — rewriting IRL rules in ARL, refactoring Java XOM classes and Spring Boot service layers, with JUnit/Mockito-based rule-testing and unit-testing frameworks validating decision tables and services in CI pipelines.
• Designed complex rule solutions (Decision Tables, Trees, IRL/ARL) with governance for change management and audit; integrated BPM with ODM, SharePoint Online, and Identity Hub via REST/SOAP APIs built on Java and Spring Boot.
• Architected and administered AWS infrastructure (EC2, S3, RDS, ELB, VPC, IAM) hosting Java/Spring Boot decision-service deployments, partnering with infrastructure teams on capacity planning, environment provisioning, and cloud migration of on-prem BPM/ODM workloads.
• Directed Azure IAM and AWS access governance, defining least-privilege roles and access policies for Spring Boot services and CI/CD pipelines deploying to cloud environments.
• Established cloud-ready deployment pipelines for Java/Spring Boot services — packaging, environment configuration, and automated promotion across Dev/QA/Production AWS environments — improving release consistency and reducing manual deployment effort.
• Led business analysis across regulatory, clinical, and HR programs — authored BRDs/FRDs, facilitated JAD sessions and BPMN modeling, maintained full requirements traceability, and served as liaison across onshore/offshore teams and 30+ stakeholders, translating business needs into Java/Spring Boot technical solutions.
• Partnered with product owners, compliance, and business stakeholders to convert regulatory and clinical requirements into architecture decisions, prioritized backlogs, and detailed technical specifications for the Java/Spring Boot engineering team.
• Ran UAT end to end with formal business sign-off; delivered training and SOPs enabling business users to self-manage compliance rules in Decision Center — direct experience driving tool adoption and enablement in a governance-heavy environment.
• Managed Acunetix DAST/SAST assessments with OWASP remediation for Java/Spring Boot services, Azure IAM access governance, and comprehensive test planning.