Browse categories
Explore
Fiverr Pro
English
$
USD
I will build production agentic ai systems with claude and AWS bedrock
Papan D
About this gig
I build production-grade agentic AI systems using Claude (Opus, Sonnet, Haiku) and AWS Bedrock. Over the past two years I've shipped multi-agent orchestrators, RAG platforms, and autonomous AIOps agents for real client workloads, not demos.
What I bring:
- Multi-agent orchestration with versioned handoffs, retry logic, and completion gates
- RAG platforms built on Bedrock Knowledge Bases and OpenSearch Serverless
- Event-driven architecture on Lambda, DynamoDB, SNS/SQS, and ECS/Fargate
- Checkpoint and recovery design for long-running agents (zero duplicate side-effects)
- Full test coverage (unit + mocked AWS integration)
Recent results:
- Cut incident MTTR by 92% (40 min to 3 min) with an autonomous remediation agent
- Built a multi-tenant RAG document Q&A platform with per-tenant data isolation
- 100% test pass rate across a 16-test suite for a multi-agent fulfillment system
I'm 4x AWS certified (Solutions Architect, ML Engineer, AI Practitioner, Cloud Practitioner) and pursuing Claude Certified Architect certification. I write clean, documented code and communicate clearly throughout the build.
Let's talk about what you're building.
Get to know Papan D
Papan D
Agentic AI, MLOps, AWS, AIOps, Blockchain
- FromIndia
- Member sinceJul 2023
Languages
English, Hindi, Bengali, Assamese
10+ years building production blockchain and AI systems. I specialize in:
▸ Agentic AI: multi-agent orchestration with Claude (Opus/Sonnet/Haiku) & AWS Bedrock
▸ Blockchain: Hyperledger Fabric, Solidity/EVM/Polygon, DAML/Canton
▸ AWS: Lambda, SageMaker, Step Function, DynamoDB, ECS/Fargate, CDK
Recent wins: cut incident MTTR by 92% with an autonomous AIOps agent, and built a cross-ledger escrow bridge settling multi-million dollar trades in under 30 seconds.
4x AWS certified. I write clean, tested code (unit + integration), document everything, and communicate clearly throughout the project.
My Portfolio
FAQ
What happens if the delivered system doesn't work as expected?
I test thoroughly before delivery (unit + integration tests), and revisions are included specifically to resolve any gaps against the agreed scope.
Do you sign NDAs?
Yes, happy to sign an NDA before discussing project details if needed.
What's your typical communication style during a project?
I send progress updates at key milestones and flag any blockers early rather than going quiet until delivery.
Can you join my team's existing codebase or repo?
Yes, I can work directly in your GitHub repo and follow your existing conventions and CI/CD setup.
How do you handle revisions?
Each tier includes a set number of revisions (listed in the package). Revisions cover refinements to the agreed scope, not new feature requests, which are handled as a separate order or add-on.
Will I get the source code?
Yes, full source code and documentation are included in every tier. Nothing is locked behind a black box.
Do I need to provide AWS access, or can you set up the environment?
Either works. I can build in your AWS account with scoped IAM access, or set up infrastructure from scratch and hand over credentials at delivery.
Which AI models do you work with?
Primarily Claude (Opus, Sonnet, Haiku) via Anthropic API and AWS Bedrock. I can also advise on other foundation models if your project calls for it.
Do you provide ongoing support after delivery?
Not included by default, but available as a separate retainer or add-on if you need it.
