← Back to Students
Shani F.
Bio
Self-taught and quick learner, highly persistent, with strong analytical and logical skills, and excellent interpersonal abilities.
Skills
Python
C++
C
Node.js
.NET
React
TypeScript
SYCL
GPU
Angular
SQL
Bootcamp Project
Extending CPU/GPU LLM inference kernels
Mentored by: Mobileye
Embedded Systems Bootcamp 2025 (Embedded)
Responsibilities:
Operator Implementation: Implemented missing SYCL operators, such as repeat_back, ABS, and ELU, with full alignment to ggml’s data structures and correct execution on both GPU and CPU.
GPU Optimization: Performed advanced GPU optimizations for the operators, resulting in a significant reduction in the number of assembly instructions and achieving a 2.57× performance improvement.
Generic Kernel Development: Developed a single generic SYCL kernel supporting a large set of operators, reducing code duplication, improving maintainability, and enabling faster backend expansion.
Upstream Integration: All implementations successfully passed code review and were fully merged into the main llama.cpp repository.
Click to enlarge
Additional Projects
Project 1 – Full-Stack Barter Exchange System (Client–Server)
A complete full-stack project enabling service-for-service exchanges without any monetary payment.
The system includes user management, posting offers, submitting requests, and matching exchanges between users.
Project 2 – Email Inbox Simulation (Offline Environment)
Background:
During the first year of my programming studies, the agenda was clear:
no internet usage at all, to train us to think independently.
There was no option to send emails, use chats, or any communication platform — only a shared file drive accessible to all students.
Because the programming tracks were spread across several floors, there was a real need to pass messages from one floor to another.
As a result, I developed an independent project: a system that simulates an email inbox.
How it worked:
For every “email address,” a file with the same name was created.
Any new message was simply written into the corresponding file.
Once the file was updated, the page re-rendered automatically so the user immediately saw the new message.
Challenges:
When several messages were sent simultaneously, they overwrote each other.
To solve this, I implemented synchronization mechanisms and locks to prevent race-condition overwrites.
After receiving approval from the program administration, the system was used by the students and provided a meaningful and practical solution — as well as a strong sense of personal accomplishment.
Technologies Used
Programming Languages & Frameworks: Angular, C#, React, Node.js
Databases: MongoDB, SQL
English Level
Working Proficiency