Hi, I'm KURISETI RAVI SRI TEJA

Machine Learning Engineer • CSE IITD@23

Kuriseti Ravi Sri Teja

About Me

I'm a cool, quick-witted guy who always loves to smile. Currently working as a Machine Learning Engineer at Eightfold AI, I specialize in Artificial Intelligence and Large-Scale Systems.

Location

Bangalore, India

Age

24 Years

Digital Twin Avatar
SYSTEM: READY
AI Cognitive Twin

Talk to My Digital Twin

Have questions about my research, professional journey, or projects? Converse directly with my digital twin to get responses in real-time.

💬

Real-Time Conversations

Ask anything about my engineering path and get precise, instant insights.

🧠

Custom Knowledge Base

Deeply aligned with my academic journey at IIT Delhi and my profesional experiences.

Education

2019 - 2023

B.Tech in Computer Science

Indian Institute of Technology, Delhi

Specialization in Artificial Intelligence. Grade: 8.64/10

2017 - 2019

Intermediate

Sri Chaitanya Junior College, Vijayawada

Grade: 10/10

2017

High School

Sri Chaitanya School, Vijayawada

Grade: 9.8/10

Experience

Nov 2025 - Present

Machine Learning Engineer

Eightfold AI, Bangalore

Building the Agentic AI Interviewer

Oct 2024 - Nov 2025

Member of Technical Staff-2

Cohesity, Bangalore

Successfully spearheaded OS migration from CentOS 7 to RHEL 9 across on-premises, multi-cloud, and hybrid environments, including comprehensive updates to SELinux policies, firewall configurations, and networking stack.

July 2023 - Sept 2024

Member of Technical Staff-1

Cohesity, Bangalore

Designed and implemented RESTful APIs for remote cluster management, enabling seamless control of IPMI users and storage devices.

July 2022 - May 2023

Undergraduate Researcher

Vision-Lab IIT Delhi

Worked on object detection using transformers.

May 2022 - July 2022

Software Development Intern

Cohesity, Bangalore

Performed backend changes to support Cohesity GCP NGCE Clusters.

Technical Skills

Publications

WACV 2024

Favoring One Among Equals - Not a Good Idea: Many-to-One Matching for Robust Transformer Based Pedestrian Detection

Authors: K.N. Ajay Shastry, K. Ravi Sri Teja, Aditya Nigam, Chetan Arora

In this paper, we investigate the reasons for lower performance of transformer-based pedestrian detection models compared to CNN-based ones. We propose a Min-cost-flow based matching algorithm that allows many-to-one matching, leading to significant improvements in miss rates across multiple datasets.

Recent Blog Posts

June 20, 2026

Graph Neural Networks: Understanding GNNs and the Problems They Solve

An intuitive and conceptual introduction to GNNs, their core message-passing mechanism, and how they solve node, edge, and graph-level problems.

Read Post