Ishaan Snehal Parikh
B.E. CSE @ RVCE | B.S. Data Science @ IIT Madras |Samsung PRISM Intern’26| Former Agentic AI Intern @ Tata AIA | Head of Operations at ACM RVCE| Co-founder & COO, Gujarat Junction Cloud Kitchen
Bengaluru
1+ year of work experience
Jun 20264 moPresent
Bengaluru, Karnataka, India
Feb 20269 mosPresent
-Currently working on the project titled "Lightweight MobileViT-Graph Network for Topological Rural Road
Extraction from Satellite Imagery" under the IAmPro 2026 Program
Dec 202511 mosPresent
Bengaluru, Karnataka, India
-Ensured the recruitment process of incoming batches was conducted successfully and transparently
-Successfully organised the ACM CTF in March 2026, and contributed as a question setter
Nov 20251 yrPresent
Bengaluru, Karnataka, India
Samsung PRISM Intern – AI Operating Systems (AIOS) Scheduler Research at Samsung Research Institute Bangalore (SRIB)
-Researched AI Operating System (AIOS) scheduling strategies for concurrent on-device AI workloads, evaluating how transformer models compete for CPU, GPU, NPU, and memory resources in heterogeneous mobile environments.
-Designed and implemented 7 scheduling algorithms (FCFS, SJF, SRTF, Round Robin, Lottery, Priority, and MLFQ), building a simulation-driven benchmarking framework using the ONNXim NPU simulator to evaluate GPT-2, BERT-Base, and OPT-125M inference workloads.
-Benchmarked 89+ concurrent AI inference tasks, analyzing latency, throughput, fairness, waiting time, resource utilization, and deadline compliance to identify trade-offs between responsiveness, QoS, and starvation prevention.
-Improved deadline satisfaction from ~1% (FCFS) to 100% (Preemptive Priority) and achieved 94.4% deadline compliance with MLFQ, while reducing lightweight inference latency from >900 ms to ~4 ms and maintaining high hardware utilization.
-Delivered data-driven recommendations for AI-aware scheduling policies balancing responsiveness, fairness, throughput, and QoS, contributing insights for future Samsung on-device AI platforms.
-Tech used: C++ · Python · ONNXim · ONNX Runtime · GPT-2 · BERT-Base · OPT-125M · AI Scheduling Algorithms · Operating Systems · CPU/GPU/NPU Simulation · Performance Benchmarking · Linux · Git
Aug 20251 yr 3 mosPresent
Bengaluru, Karnataka, India
-Engineered a production-ready Agentic AI e-KYC chatbot orchestrating 5+ document types using LangGraph-based, state-controlled multi-agent workflows, OCR, and retrieval-augmented classification, reducing manual verification by ~60–70% and onboarding time by ~40–50%.
-Implemented a hierarchical memory architecture (L1/L2/L3) using Redis and vector databases, enabling the e-KYC chatbot to remember user details across sessions, reduce KYC decision time by ~35–40%, integrate smoothly with Lead-to-Issuance (L2I) workflows, and improve L2I efficiency.
-Delivered the system end-to-end within a 6-week internship timeline, following clean coding practices and structured reviews.
May 20251 yr 6 mosPresent
Bengaluru, Karnataka, India
-Member of the Open Source and Development Team at Google Developers Group, RVCE
-Successfully conducted the recruitment for the incoming batches
-Assisted the organising committee in the Vision2047 Hackathon in registrations and event management
Dec 20241 yr 1 moDec 2025
Bengaluru, Karnataka, India
-Successfully organized the Tech Tank 2025 from April 7-9 2025 and helped the organizing committee as a volunteer.
Ishaan’s AI Stack
Education
Bachelor of Science - BS, Data Science and Applications
Indian Institute of Technology, Madras
20253 yrs2028
Bachelor of Engineering - BE, Computer Science and Engineering
RV College of Engineering
20244 yrs2028
Join Ishaan and 10k+ other professionals on
Tal is the app where bosses meet engineers & builders directly, without middlemen
Download Tal Boss