Ravi Kant Sharma profile photo 1 on Tal

Ravi Kant Sharma

SDE@Amdocs | GenAI | Agentic AI | Java | Spring Boot | Microservices | MCP | RAG | MERN | SQL | Kafka | Cloud | DTU 24

Gurugram

2+ years of work experience

Amdocs
Aug 20242 yrs 1 moPresent

Gurugram, Haryana, India

InnOVerse (Agentic Crowd) — GenAI Team Technologies: React, Node.js, Express.js, FastAPI, PostgreSQL, Redis, Docker, Podman, Nginx, RBAC/ABAC, GitLab CI/CD • Re-architected an internal GenAI asset platform (MCP servers, rules, AI tools, skills) into a multi-tenant SaaS product for 4 telecom clients via a shared-core + per-client-delivery Feature-Sliced architecture. • Designed a centralized RBAC / Policy Decision Point (ABAC-lite) — 20 permissions, 10 roles, 3 enforcement modes — and led a zero-downtime migration off a legacy role enum across backend, JWT, and frontend. • Built a FastAPI semantic / vector-search (RAG) microservice (7 endpoints, top-K ranking) with 15s timeouts and graceful fallback, powering retrieval, similarity, and duplicate detection. • Built a broker-free, PostgreSQL-backed notification system (33 typed notifications, 8 domains, 2 channels) with O(1) broadcast fan-out, removing a message-broker dependency. • Implemented enterprise auth (Azure AD SSO / MSAL, NTLM, LDAP) with a Redis-backed session store; shipped via GitLab CI/CD as separate Podman images behind an Nginx TLS edge. Real-Time Billing & SKY Fallout Management (AT&T, USA) Technologies: Java, Spring Boot, React, Kafka, Couchbase, PostgreSQL, Jenkins, OpenShift, Postman • Enhanced the fallout-handling stack (Java, Spring Boot, Jenkins, JESI, Kafka, PostgreSQL, React) into a reliable, scalable solution integrating multiple workspaces, migrating 1,000,000+ customers seamlessly. • Implemented caching across three Spring Boot microservices using Couchbase (15-min TTL) to speed up API lookups. • Enhanced the billing simulator to support amend/cancel orders via dedicated Kafka consumers and producers. • Built a new fallout-management workspace for customer migration, auto-updating dashboards on Kafka events with custom auto/manual ticket resolution. • Built EnvStatus, an MS Teams bot for environment status and pod versions (OpenShift).
Echio
Dec 20232 mosJan 2024

Remote

QuickReel
Jan 20234 mosApr 2023

New Delhi, Delhi, India

At QuickReel, I built a full-stack AI-based video editing platform that automatically generated social media-ready clips. On the backend, I developed REST APIs in Node.js/Express, integrated with AWS S3 for video storage, and used MongoDB as the database. I containerized the backend using Docker, ensuring consistent deployments and environment isolation. I optimized heavy video processing pipelines by leveraging OpenCV and efficient parallelization techniques, reducing processing time from 30 minutes to 45 seconds while also cutting bandwidth usage by 85%. On the frontend, I implemented a React + Redux UI with a custom-built web-based video player, supporting preview and playback of processed clips. Additionally, I integrated Stripe APIs for subscription management (renewals, cancellations, billing), and added OAuth-based authentication for secure access.
GitHub
@rksharma007
LeetCode
@rksharma007
Personal portfoliorksharma.in

Ravi’s AI Stack

claude code
claude code
cursor
cursor
langchain
langchain
langgraph
langgraph
mcp
mcp
claudecode
claudecode

Education

Delhi Technological University (Formerly DCE)

Bachelor of Technology, Computer Software Engineering

Delhi Technological University (Formerly DCE)

20204 yrs2024

Join Ravi and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Ravi Kant Sharma on Tal

Headline

SDE@Amdocs | GenAI | Agentic AI | Java | Spring Boot | Microservices | MCP | RAG | MERN | SQL | Kafka | Cloud | DTU 24

About

Software Engineer with hands-on experience in building scalable distributed systems, event-driven architectures, and full-stack applications. Currently working on an Agentic AI platform, Agentic Crowd (InnOVerse), for managing organization-wide GenAI Assets (MCP, Skills, Rules, Personas) at Amdocs. Amdocs utilizes my skills in developing and optimizing Spring Boot microservices for large-scale telecom systems. I contribute to event-driven workflows with Kafka, enhance fallout management with SQL and NoSQL databases, and automate billing processes using YAML configurations. My responsibilities also include containerizing and deploying services on Docker and managing workloads on OpenShift/Kubernetes.

Location

Gurugram

AI stack

claude code

cursor

langchain

langgraph

mcp

claudecode

Work experience

Software Developer at Amdocs, Aug 2024 - Present · 2 yrs 1 mo, Gurugram, Haryana, India, InnOVerse (Agentic Crowd) — GenAI Team Technologies: React, Node.js, Express.js, FastAPI, PostgreSQL, Redis, Docker, Podman, Nginx, RBAC/ABAC, GitLab CI/CD • Re-architected an internal GenAI asset platform (MCP servers, rules, AI tools, skills) into a multi-tenant SaaS product for 4 telecom clients via a shared-core + per-client-delivery Feature-Sliced architecture. • Designed a centralized RBAC / Policy Decision Point (ABAC-lite) — 20 permissions, 10 roles, 3 enforcement modes — and led a zero-downtime migration off a legacy role enum across backend, JWT, and frontend. • Built a FastAPI semantic / vector-search (RAG) microservice (7 endpoints, top-K ranking) with 15s timeouts and graceful fallback, powering retrieval, similarity, and duplicate detection. • Built a broker-free, PostgreSQL-backed notification system (33 typed notifications, 8 domains, 2 channels) with O(1) broadcast fan-out, removing a message-broker dependency. • Implemented enterprise auth (Azure AD SSO / MSAL, NTLM, LDAP) with a Redis-backed session store; shipped via GitLab CI/CD as separate Podman images behind an Nginx TLS edge. Real-Time Billing & SKY Fallout Management (AT&T, USA) Technologies: Java, Spring Boot, React, Kafka, Couchbase, PostgreSQL, Jenkins, OpenShift, Postman • Enhanced the fallout-handling stack (Java, Spring Boot, Jenkins, JESI, Kafka, PostgreSQL, React) into a reliable, scalable solution integrating multiple workspaces, migrating 1,000,000+ customers seamlessly. • Implemented caching across three Spring Boot microservices using Couchbase (15-min TTL) to speed up API lookups. • Enhanced the billing simulator to support amend/cancel orders via dedicated Kafka consumers and producers. • Built a new fallout-management workspace for customer migration, auto-updating dashboards on Kafka events with custom auto/manual ticket resolution. • Built EnvStatus, an MS Teams bot for environment status and pod versions (OpenShift).

Software Developer Intern at Echio, Dec 2023 - Jan 2024 · 2 mos, Remote

Software Developer Intern at QuickReel, Jan 2023 - Apr 2023 · 4 mos, New Delhi, Delhi, India, At QuickReel, I built a full-stack AI-based video editing platform that automatically generated social media-ready clips. On the backend, I developed REST APIs in Node.js/Express, integrated with AWS S3 for video storage, and used MongoDB as the database. I containerized the backend using Docker, ensuring consistent deployments and environment isolation. I optimized heavy video processing pipelines by leveraging OpenCV and efficient parallelization techniques, reducing processing time from 30 minutes to 45 seconds while also cutting bandwidth usage by 85%. On the frontend, I implemented a React + Redux UI with a custom-built web-based video player, supporting preview and playback of processed clips. Additionally, I integrated Stripe APIs for subscription management (renewals, cancellations, billing), and added OAuth-based authentication for secure access.

Education

Bachelor of Technology, Computer Software Engineering, Delhi Technological University (Formerly DCE)

Public links

github: https://github.com/rksharma007

leetcode: https://leetcode.com/rksharma007

rksharma.in: https://rksharma.in/