About Plavaga

Built From the Inside. That's How We Know What Works.

We've shipped AI across hospitality, health tech, real estate, and fintech. The work holds up because we build it ourselves — architecture to production.

Our journey

How We Got Here

From infrastructure shop to AI diagnostic practice.

2010

Production Software

Full-stack platforms for business networking, real estate, e-commerce, ed-tech, enterprise software. Clients that stayed 5–6 years.

2013

Deep into AWS

Cloud Architecture, DevOps, Legacy Modernization, Security. The infrastructure foundation for everything that followed.

2016

IoT Platform & Predictive Analytics

Built an IoT PaaS with predictive maintenance analytics and event-driven data ingestion pipelines. ML in production before the LLM wave.

2019

Long-term Technical Partnership

Multi-year engagements owning infrastructure, security architecture, and engineering roadmaps for product companies.

2023

Production AI

Architected and then led engineering. Shipped conversational commerce (hospitality), medical AI (health tech), content moderation (consumer wellness), document intelligence (real estate). Real users, real production incidents.

See the AI systems →
Now

Diagnostic Practice

Same pattern across every industry: the model ships, the operational layer doesn’t follow. Built a practice around diagnosing that gap.

See diagnostic services →

AI-native isn’t a marketing claim. It’s how we build.

Including this website.

The Team

Rishi Choudhary

Rishi Choudhary

AI Implementation & Technical Architecture

Shipped AI across hospitality (conversational commerce over WhatsApp), health tech (FHIR-based medical AI), and consumer wellness (AI content moderation). Some made it to production users. Some didn’t survive the startup. All taught him where AI breaks — and why. Advises QuietGrowth (fintech robo-advisory) on AI architecture in a regulated context. 20+ years across Infosys, EMC, and HP before focusing on AI implementation. When RAG failures or monetization gaps surface, Rishi has usually seen the pattern before.

Azmi Ahmad

Azmi Ahmad

AWS Infrastructure & AI Integration

Over a decade on AWS — infrastructure, security architecture, DevOps. Certified Solutions Architect. Outsourced CTO at Aarca Research since 2019, now spanning infrastructure, security, and AI tool integrations into their product roadmap. He brings both the infrastructure depth that keeps AI systems running in production and hands-on experience integrating AI into a real health tech product. When margin bleed or security gaps get diagnosed, Azmi is the reason the fix actually deploys.

When projects need more hands, we scale with vetted domain specialists — AI/ML engineers, security architects, frontend developers. The diagnostic scopes the work; scaling happens at the build phase.

Client voices

From people we’ve worked with

Plavaga has been our technology backbone since inception. From cloud architecture to AI-powered diagnostics, they brought the technical depth that a health-tech company needs but can rarely find in one team.

Gayathri Choda

Founder, Aarca Research

Plavaga has taken the initiative of meticulously understanding what we wanted and delivered beyond par.

Prakash Sethuraman

Managing Partner, Enterprise Blueprints

Plavaga understood the business requirements and made excellent suggestions for improving the customer experience.

Christina Ravaglia

Senior Director of Product Management, Satmetrix

We have been working with Plavaga for over 5 years. They displayed tremendous maturity in understanding the problem we were attempting to solve.

Ravi Shankar

Founder Director, eMyPA

Your experience and guidance has been the perfect match for our enthusiasm. Working with you all has been a pleasure.

Nia

Founder, Taglr

What We Believe

Diagnose, don’t pitch.

Every call starts with your problem — what’s hurting, what’s worth looking at, and a straight answer. Half the calls end with ‘you don’t need us for this.’

See our services →

Senior engineers. No layers.

The people who diagnose are the people who build. No handoffs, no ‘let me check with the team.’ Architecture decisions happen in the room, not after three rounds of telephone.

AI-native, not AI-adjacent.

Claude Code, LLM agents — in the daily workflow, not on the website for show. AI in production is understood here because it’s built with every day.

Start bounded, not open-ended.

Every engagement begins with a scoped diagnostic. Fixed timeline, fixed cost, written deliverable. You see the findings before you decide what (if anything) to build next.

Technology Stack

Experience spanning AI, cloud architecture, IoT, and full-stack product development — from sensor data pipelines to LLM-powered systems.

AI & Intelligence

AI & ML

LangChainLangGraphpgvectorNeo4jElasticsearchPineconeOpenSearchOpenAI APIAnthropic APIGoogle Gemini APIAWS BedrockMCP (Model Context Protocol)ACP (Agentic Commerce Protocol)LangfuseLiteLLMAI4Bharat IndicTrans2Ollaman8n

AI Billing & Metering

StiggStripe Billing

AI Dev Tools

CursorClaude CodeOpenRouter

Cloud & Infrastructure

AWS Compute

ECSFargateLambdaStep FunctionsEC2EKSSageMaker

AWS Storage & Data

S3RDS (PostgreSQL)DynamoDBElastiCacheOpenSearch

AWS Security

KMSIAMSecrets ManagerWAFVPCCloudTrail

AWS Networking

CloudFrontAPI GatewayALBRoute 53SQSSNSEventBridgeKinesis

Infrastructure as Code

TerraformAWS CDKCloudFormation

Containers & Orchestration

DockerKubernetesHelm

CI/CD

GitHub ActionsArgoCD

Development

Languages

TypeScriptPythonJava (Spring Boot)Node.js

Frontend

ReactNext.js

Observability

CloudWatchX-RayDatadogGrafanaOpenTelemetry

Edge

Hardware/IoT

ESP32MQTTBLECoAP

Now You Know Who Does the Work.

Tell us where you are with AI. Honest feedback on what makes sense — and what doesn't. 30 minutes, no pitch.