Prateek Kumar

Empowering Next-Gen ML & AI Platforms

Available for projects
Get To Know Me

👤 About Me

👋Hello, I'm Prateek Kumar

A passionate ML Platform Engineer and  Cloud Native enthusiast with 5+ years of experience crafting scalable ML platforms and cloud-native solutions.

At AMD, I lead efforts within the AI Group team to develop unified, high-performance AI software solutions that serve millions worldwide. I thrive on tackling hyper-scale challenges, constantly pushing the boundaries of what's possible in the ML and cloud-native space.

Whether it's designing scalable architectures or optimizing performance, I'm dedicated to delivering robust, efficient systems that keep pace with today's dynamic tech landscape.

Core Technologies

PythonKubernetesDockerAWSElasticsearchPrometheusGrafanaKafkaPostgreSQLTerraformKubeflowArgoCD

Let's Connect

prateekkumar1709@gmail.comLinkedInGitHubBlog
New York, USA

Newsletter

Subscribe to get the latest updates on ML platforms, cloud-native tech, and more.

Subscribe on LinkedIn
Professional Journey

💼 Experience

My professional journey and the impact I've made across different domains

AMD AI Group logo
June 2025 — Present

Senior Staff Software Engineer

AMD AI Group

Working with the AI Group team at AMD, driving innovation in AI infrastructure and platform engineering for next-generation AI solutions.

Key Achievements

Developing cutting-edge AI infrastructure solutions that power AMD's next-generation AI initiatives
Building scalable AI platforms optimized for AMD's latest hardware capabilities, delivering unprecedented performance
Collaborating with cross-functional teams to architect and implement innovative AI infrastructure solutions
Contributing to high-performance AI software development, focusing on system optimization and scalability

Technologies & Skills

AI InfrastructurePlatform EngineeringSystem ArchitecturePerformance OptimizationCloud Native
INTENT HQ logo
May 2023 — Present

Software Engineer (ML Platform)

INTENT HQ

Building and scaling ML platforms at a 100+ employee venture-backed AI startup in New York, NY.

Key Achievements

Built scalable data processing platform integrating Logstash and OpenSearch, processing 5TB+ daily ML model outputs from S3 and Kafka streams for real-time customer insights
Architected GitOps-based MLOps platform using Dagster and ArgoCD, reducing the ML pipeline development cycle by 45% while ensuring security compliance
Implemented custom Kubernetes-native secret management solution using HashiCorp Vault and External Secrets Operator, enabling secure deployment across 200+ microservices
Established production infrastructure using Helm and Kustomize, managing 50+ microservices across 4 environments with custom operators, resulting in 65% reduced deployment complexity and 90% configuration reuse
Orchestrated Kubernetes-based Logstash pipelines with custom Ruby filters for ML data transformation, processing 5TB+ daily training data with intelligent routing to OpenSearch clusters
Led PoC initiative to modernise observability stack using Grafana Beyla and eBPF, eliminating sidecar deployments across 100+ services while reducing monitoring overhead by 35%
Mentored junior engineers in cloud-native technologies and best practices, fostering a culture of knowledge sharing and continuous learning
Led cross-functional initiatives to standardize development practices and improve system reliability across teams

Technologies & Skills

KubernetesDockerAWSLogstashOpenSearchKafkaDagsterArgoCDHashiCorp VaultHelmKustomizeGrafanaeBPFTeam LeadershipTechnical Mentoring
PIVOTREE logo
July 2017 — July 2021

Senior Software Engineer

PIVOTREE

Led development of enterprise monitoring solutions and cloud-native platforms in Bengaluru, India.

Key Achievements

Spearheaded the development of YANTRA Control Tower, a proprietary monitoring solution for order management systems, driving $1M+ revenue through real-time operational insights
Led team of 5 developers in designing and implementing a comprehensive observability ecosystem reducing MTTR by 30% and improving application performance by 25%
Orchestrated migration to cloud-native Kubernetes platform with vendor-agnostic architecture, reducing client onboarding by 60% and scaling to 30+ enterprise clients
Led technical recruitment initiatives and conducted interviews for software engineering roles, successfully hiring 20+ engineers while maintaining high talent standards and team culture alignment
Established engineering best practices and coding standards, resulting in 40% reduction in production incidents
Mentored junior developers and conducted regular knowledge sharing sessions, improving team productivity by 35%

Technologies & Skills

KubernetesDockerAWSPrometheusGrafanaELK StackMicroservicesCI/CDTerraformAnsiblePythonJavaTeam LeadershipTechnical MentoringEngineering Management
</>Open Source

</> Open Source Footprint

Giving back to the community through code and leadership.

OpenGitOps logo

OpenGitOps

  • OpenGitOps Project Member: Active member of the OpenGitOps project, contributing to the advancement of GitOps practices and standards.
  • Whitepaper Lead: Leading initiatives on writing whitepapers about Secrets Management in GitOps and GitOps Patterns.

✨ Current Whitepaper Contributions ✨

CNCF logo
In Progress
Challenges in Scheduling Cloud Native AI Workloads
In Progress
Foundations & Framework for Sustainable AI in Cloud-Native Environments
In Progress
CNCF Cloud Native AI White Paper

🚀 Recent Pull Requests 🚀

Open Feature logo

Open Feature

Feature
Use docker buildx in CI to build containers for other architectures including ARM
Open Model Initiative logo

Open Model Initiative

Feature
Add DCO sign off pre-commit hook
Feature
Add smoke tests api db
Bug Fix
Fix reset-postgres task volume name
Enhancement
Enhancement/flake8 pipeline check
Open Telemetry Collector logo

Open Telemetry Collector

Enhancement
Add mdatagen replace option to check-contrib target
External Secrets Operator logo

External Secrets Operator

Feature
Add values.schema.json generation to Helm chart
Maintenance
Bump helm-docs image version in Makefile
Tech Arsenal

🛠️ Technology Stack

Technologies I use to bring ideas to life

Programming Languages

Python

SQL

Shell Scripting logo

Shell Scripting

Java

Cloud

AWS

Docker

Kubernetes

Istio logo

Istio

Helm

Crossplane logo

Crossplane

ArgoCD logo

ArgoCD

KServe logo

KServe

Kubeflow logo

Kubeflow

Databases

Elasticsearch

PostgreSQL

DynamoDB logo

DynamoDB

Oracle DB

Vector Databases

Redis

Observability & Monitoring

Logstash

OpenSearch logo

OpenSearch

Kibana

Prometheus

Grafana

Tools & Platforms

Kafka

Git

CI/CD

Dagster logo

Dagster

HashiCorp Vault

Terraform

Operating Systems

Linux

Windows

Contact

🤝 Let's Work Together

Ready to bring your ideas to life? I'm always excited to work on interesting projects and collaborate with amazing people. Let's create something extraordinary together.

Or reach out directly: