s
sumit_d_patel

Sumit Patel

@sumit_d_patel

You never know, What

Índia
Inglês, Hindi, Marata
Algumas informações são exibidas no idioma inglês.
Sumit Patel está indisponível até May 14, 2027

Você ainda pode contatar o prestador.

Sobre mim
Backend and DevOps engineer experienced in building and deploying scalable applications. I work primarily with python (Flask/FastAPI), NextJS, and NodeJS, Managing deployments using docker, Nginx and Linux servers. My projects include SaaS backends, API Automations tools, and cloud infrastructure setups with AWS and OCI. I focus on reliability, clean architecture, and production grade efficiency.... Saiba mais

Habilidades

s
sumit_d_patel
Sumit Patel
offline • 

Experiência profissional

Backend & DevOps Engineer

Kumari AI • Período integral

Jul 2024 - Aug 2025 • 1 yr 1 mo

• Designing and building backend infrastructure that supports orchestration of multiple AI models/APIs. • Implementing routing logic: building systems to select the appropriate LLM/tool per request, balancing tradeoffs (cost, latency, quality). • Setting up DevOps pipelines: CI/CD, deployment, monitoring, scaling, ensuring the system handles many concurrent requests reliably and scales horizontally as usage grows • Integration of AI services: managing credentials, request routing to external model providers (e.g. OpenAI, etc.), handling failures/fallbacks, logging/telemetry, quality tracking

DevOps Intern

PeerXP Services • Período integral

Feb 2023 - Aug 2023 • 6 mos

• Engineered and implemented comprehensive monitoring solutions using Grafana and Prometheus, resulting in improved system visibility and reduced incident response time from hours to minutes • Developed and optimized Docker containers integrated into client automation pipelines, reducing deployment time and improving resource usage. • Automated CI/CD workflows on AWS using GitLab CI, reducing manual deployment steps and maintained system configurations via Ansible and Terraform • Authored 1000+ self-hosted, wiki-style setup guides, used by new hires and external developers, reducing onboarding and support queries