Shreyas S profile photo 1 on Tal

Shreyas S

Software Engineer | GSoC’24

Bengaluru

3+ years of work experience

Stealth
Dec 20259 mosPresent

United States

Founding engineer building core infrastructure for a headless CMS platform. - Own multiple critical systems spanning scripting engine, routing, blocks, deployments. - Designed the expression language that powers dynamic content relationships - Built caching architecture and delivered 4x performance improvements - mass ~30% of platform functionality; - Day 1 contributor
Puch AI
Jul 20253 mosSep 2025

Bengaluru

Optimized real-time AI speech and multilingual systems at scale for millions of Indian users. - Parallelized STT on H100 GPUs with CUDA streams ; 18% latency reduction - Scaled MCP architecture from single to multi-server per user - Multilingual pipeline for 11 Indic languages with robust script detection & fallbacks - Integrated audio generation within existing in-house text-to-video workflows
ScoreTravel
Mar 20254 mosJun 2025

California, United States

Built data quality infrastructure to reduce noise and cost in high-throughput ML pipelines. - Deduplication microservices using text similarity and image processing benchmarks - R&D experiments tracked via WandB; validated against VLMs for approach effectiveness
Emory University Department of Biomedical Informatics
Sep 20244 mosDec 2024

Atlanta, Georgia, United States

Interpretable deep learning for large scale EEG modeling. ( extended on top of GSoC work) - DeepLIFT attribution pipeline to identify key brain signal features - Vector quantization training with multi-node synchronization - Visualization tools (heatmaps, time-resolved plots) for channel-wise analysis
AIR CENTER VITAP
Mar 20232 yrs 9 mosNov 2025

On-site

Generative AI & computer vision research across image synthesis, clustering, and document understanding. - Stable Diffusion inpainting + LoRA fine-tuning for context-aware image editing - DBScan-based multi-face clustering for identity preservation at scale - Noval SOTA OCR pipeline for artifact extraction from mechanical drawings
GitHub
@zhreyu
LeetCode
@zhreyu
Personal portfoliozhreyu.github.io

Shreyas’s AI Stack

anthropic apis
anthropic apis
claude code
claude code
claude
claude
codex
codex
cursor
cursor
opencode
opencode

Education

Vellore Institute of Technology

Master of Technology, Computer Science and Engineering ( 5 Year Integrated )

Vellore Institute of Technology

20225 yrs2027
Kendriya Vidyalaya Noida, New Delhi

Kendriya Vidyalaya Noida, New Delhi

Join Shreyas and 10k+ other professionals on

Tal Boss

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

Download Tal Boss

Shreyas S on Tal

Headline

Software Engineer | GSoC’24

About

Ping : shreyas.s@outlook.in Work : https://zhreyu.github.io/work If you're building something and think I can help, my DMs are always open :) subtle things I don’t usually mention: built a SaaS generating $1K+ in sales Finalist, Adobe India Hackathon 2025 (top 7 / 260k+ participants) Hacksquad 2023 (top 50 winning teams / 2000+ teams) won multiple college-level hackathons polyglot (7 languages, grew up across north and south India in a multilingual household)

Location

Bengaluru

AI stack

anthropic apis

claude code

claude

codex

cursor

opencode

Work experience

Software Engineer at Stealth, Dec 2025 - Present · 9 mos, United States, Founding engineer building core infrastructure for a headless CMS platform. - Own multiple critical systems spanning scripting engine, routing, blocks, deployments. - Designed the expression language that powers dynamic content relationships - Built caching architecture and delivered 4x performance improvements - mass ~30% of platform functionality; - Day 1 contributor

AI Engineering Intern at Puch AI, Jul 2025 - Sep 2025 · 3 mos, Bengaluru, Optimized real-time AI speech and multilingual systems at scale for millions of Indian users. - Parallelized STT on H100 GPUs with CUDA streams ; 18% latency reduction - Scaled MCP architecture from single to multi-server per user - Multilingual pipeline for 11 Indic languages with robust script detection & fallbacks - Integrated audio generation within existing in-house text-to-video workflows

ML Intern at ScoreTravel, Mar 2025 - Jun 2025 · 4 mos, California, United States, Built data quality infrastructure to reduce noise and cost in high-throughput ML pipelines. - Deduplication microservices using text similarity and image processing benchmarks - R&D experiments tracked via WandB; validated against VLMs for approach effectiveness

Student Researcher at Emory University Department of Biomedical Informatics, Sep 2024 - Dec 2024 · 4 mos, Atlanta, Georgia, United States, Interpretable deep learning for large scale EEG modeling. ( extended on top of GSoC work) - DeepLIFT attribution pipeline to identify key brain signal features - Vector quantization training with multi-node synchronization - Visualization tools (heatmaps, time-resolved plots) for channel-wise analysis

Student Researcher ( University Lab ) at AIR CENTER VITAP, Mar 2023 - Nov 2025 · 2 yrs 9 mos, On-site, Generative AI & computer vision research across image synthesis, clustering, and document understanding. - Stable Diffusion inpainting + LoRA fine-tuning for context-aware image editing - DBScan-based multi-face clustering for identity preservation at scale - Noval SOTA OCR pipeline for artifact extraction from mechanical drawings

Education

Master of Technology, Computer Science and Engineering ( 5 Year Integrated ), Vellore Institute of Technology

Kendriya Vidyalaya Noida, New Delhi

Public links

github: https://github.com/zhreyu

leetcode: https://leetcode.com/zhreyu

zhreyu.github.io: https://zhreyu.github.io/