AI engineer · full-stack · ships end-to-endIndia · remote (US / EU hours)

I turn messy ideas
into systems that ship.

I don't need to be told what to build. Show me the problem, I'll figure out the rest.

View my work

oss prs merged

25+ OSS PRs Merged. Across open source repos like Docker Agent, Kubernetes, Dify and more

agents shipped

7+ Agents Shipped. Orchestration, retrieval, and eval stacks shipped to production
How I think

Wired to think
like a builder.

01
Full-stack ownership
Early-stage means no specialists. Infra, agents, APIs, demos, I've owned all of it, sometimes in the same week.
02
Speed with judgment
Fast is useless if the next engineer can't understand what you built. I ship quickly and leave things readable. That's the actual bar.
03
Integration-first thinking
LLMs in notebooks aren't products. I've wired them to Jira, Outlook, WhatsApp, Shopify. The boring integration work is where most demos die. I don't let that happen.
04
Product sense
I've sat close enough to founders to know the difference between what's technically interesting and what actually needs to ship this week. That gap is where bad engineers waste your time.
Work · Internship & Projects

Things I've actually built

Scroll · drag
01
MULTI-AGENTEKSTERRAFORMINTERNSHIP

SWARM

Enterprise multi-agent system built for real workloads. Every investigation gets its own sandboxed EKS pod, with a YAML-driven abstraction so any new multi-agent workflow is a config change.

Google ADKLangGraphTerraformEKSKubernetes
Telecom AI startup · Enterprise automation infra
Read case study
02
RAGAGENT INFRAHYBRID SEARCHINTERNSHIP

RETRIEVAL CORE

The retrieval layer behind a telecom AI agent. Custom ingestion per content type, LLM-enriched embeddings at index time, and hybrid retrieval with reranking on the shortlist

QdrantFastAPICohereBM25+RRFMinIO
Telecom AI startup · Agent-facing knowledge layer
Read case study
03
AI PRODUCTWHATSAPP100K+ SCALEINTERNSHIP

AutoCLM

Autonomous WhatsApp CLM for telecom operators. A three-layer architecture separates strategy, planning, and execution so no agent ever sees more than it needs to.

LangGraphWhatsApp APIReactFastAPIMongoDB
Telecom AI startup · Built for client pitch, demoed live
Read case study
Personal projects
04
AI PIPELINENEXT.JSPERSONAL

Folio

AI-native graphic generation engine. A 3-agent pipeline picks a template, fills content slots, and critiques the rendered output before you see it.

Google ADKFastAPINext.jsTypeScriptPython
Personal · May 2026
Email about this project
05
Go · MicroservicesPersonal

Go gRPC Microservices

Built a distributed system in Go to learn gRPC and GraphQL properly — not a tutorial clone, but something with real service boundaries, protobuf contracts, and an API gateway. The kind of project you build when you want to understand the seams.

GogRPCGraphQLProtobuf
Personal · distributed systems study
Email about this project
More experiments

Not everything fits a card. If something here overlaps what you are building, ask — I will point you to the closest parallel.

WebRTC object detectionMicro ADK frameworkPDF extractorFluxShare · RakshaChakra
Email about other builds
Stack

Everything I've
shipped with

LangGraphGoogle ADKKubernetesTerraformGoPythonTypeScriptFastAPILangGraphGoogle ADKKubernetesTerraformGoPythonTypeScriptFastAPILangGraphGoogle ADKKubernetesTerraformGoPythonTypeScriptFastAPI
QdrantBM25+RRFCohere RerankOpenRouterLiteLLMAWS EKSMinIOPostgresQdrantBM25+RRFCohere RerankOpenRouterLiteLLMAWS EKSMinIOPostgresQdrantBM25+RRFCohere RerankOpenRouterLiteLLMAWS EKSMinIOPostgres
ReactgRPCGraphQLWebRTCJira WebhooksSvixAgentMailWhatsApp APIReactgRPCGraphQLWebRTCJira WebhooksSvixAgentMailWhatsApp APIReactgRPCGraphQLWebRTCJira WebhooksSvixAgentMailWhatsApp API
GitHub Contributions
Fetching GitHub data
Open source

I contribute to
things that matter.

Merged PRs on production codebases. Features, fixes, and infra changes reviewed by engineers who don't know me.

12+ merged PRs

The official AI agent runtime by Docker Engineering. Contributions across core developer tooling.

  • Background job management for shell toolset (run, list, view)
  • Filesystem tools that respect .gitignore rules across all operations
  • Auto directory creation in write_file for nested paths

Fixed a silent proxy-mode bug in the K8s operator. Added OTel trace propagation to proxy.

MCPSECURITY

Replaced 100+ lines of manual dict validation with Pydantic v2 discriminated unions.

AIPYTHON

Consolidated duplicated TypeScript shims across fixtures and examples into a single shared source of truth.

TYPESCRIPTCLOUDFLARE

Added feature gate support to the VPA local deploy script.

KUBERNETES
Contact
Let'sbuildsomethingreal.

If you are moving fast and need someone who can own hard technical problems without hand holding. My inbox is open.

Availability snapshot
StatusAvailable now
TypeFull-time · Contract
LocationIndia / Remote
NoticeImmediate
Ideal team size2–20 people
Email me