Ansh R. Savadatti profile photo 1 on Tal

Ansh R. Savadatti

SDE-1 @HashedIn By Deloitte | CSE UnderGrad at KLE Tech.

Bengaluru

2+ years of work experience

HashedIn by Deloitte
Apr 20266 mosPresent

Data Engineer
HashedIn by Deloitte
Jan 20264 mosApr 2026

Ultimez Technology Pvt Ltd
May 20252 mosJun 2025

Remote

NVIDIA
Sep 20241 yr 5 mosJan 2026

Santa Clara County, California, United States

Developed a multimodal transformer-based image comparison estimation system under the NVIDIA AI Technology Center, generating structured observations and computing semantic difference scores between image pairs. Also implemented demosaicing pipelines to convert RCCB sensor data into RGB images, improving visual reconstruction.
Code Club KLETU
Jan 20242 yrsDec 2025

Hubli, Karnataka, India

Ansh’s AI Stack

claude code
claude code
claude
claude
github copilot
github copilot
langchain
langchain
langgraph
langgraph
ChatGPT
ChatGPT

Education

KLE Technological University - Hubballi (India)

Bachelor of Engineering, Computer Science and Engineering

KLE Technological University - Hubballi (India)

20224 yrs2026

Join Ansh and 10k+ other professionals on

Tal Boss

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

Download Tal Boss

Ansh R. Savadatti on Tal

Headline

SDE-1 @HashedIn By Deloitte | CSE UnderGrad at KLE Tech.

About

Software Engineer with hands-on experience in multimodal AI systems and scalable backend engineering. I’ve worked on both AI systems and full-stack engineering, giving me a strong foundation across the stack. Recently, I worked as a Project Intern with NVIDIA under the NVIDIA AI Technology Center, where I developed a multimodal transformer-based system to compare images using structured observations and semantic similarity scoring. I also worked on RCCB to RGB demosaicing using deep learning models optimized with hybrid loss functions (L1 + SSIM). At HashedIn by Deloitte, I built scalable backend systems and full-stack applications using FastAPI, Python, and Angular. I also worked on Agentic AI systems using RAG, vector databases, MCP, and real-time communication via WebSockets and SSE. I enjoy solving complex problems and have solved 500+ problems across platforms like LeetCode, Codeforces, and CodeChef, strengthening my algorithmic thinking and system design skills. I’m particularly interested in: - Multimodal AI & Computer Vision - Backend systems & scalable architectures - AI-powered applications (RAG, agents, MCP, intelligent pipelines) Currently, I’m focused on building impactful systems and continuously improving my skills in AI and software engineering.

Location

Bengaluru

AI stack

claude code

claude

github copilot

langchain

langgraph

ChatGPT

Work experience

Software Engineer at HashedIn by Deloitte, Apr 2026 - Present · 6 mos, Data Engineer

SDE Intern at HashedIn by Deloitte, Jan 2026 - Apr 2026 · 4 mos

Full-stack Developer at Ultimez Technology Pvt Ltd, May 2025 - Jun 2025 · 2 mos, Remote

Project Engineering Intern at NVIDIA, Sep 2024 - Jan 2026 · 1 yr 5 mos, Santa Clara County, California, United States, Developed a multimodal transformer-based image comparison estimation system under the NVIDIA AI Technology Center, generating structured observations and computing semantic difference scores between image pairs. Also implemented demosaicing pipelines to convert RCCB sensor data into RGB images, improving visual reconstruction.

Co-Founder at Code Club KLETU, Jan 2024 - Dec 2025 · 2 yrs, Hubli, Karnataka, India

Education

Bachelor of Engineering, Computer Science and Engineering, KLE Technological University - Hubballi (India)

Ansh R. Savadatti - Software Engineer in Bengaluru | Tal