AI Tooling - Systems - Performance - Data EngineeringMS CS @ Case Western Reserve University
Sri Satya Sai Immani
I build practical AI systems and developer-friendly tooling: LLM integrations, retrieval workflows, reliable APIs, and performance-aware pipelines.
Library chatbot impact
30% workload drop
MCQ agent impact
~60% time saved
GPU sorting throughput
344.6M elems/s
Tech stack
.svg){kind=small}
About
SS
Graduate CS engineer focused on AI systems and performance work: LLM integration, RAG-style workflows, and GPU/CPU benchmarking, built to be measurable and reliable.
LLM IntegrationData + APIsCUDA / Benchmarking
Featured Projects
See allNovelTease
Multimodal generation + FFmpeg video synthesis behind a production-grade FastAPI backend.
Hybrid Bucket-Radix GPU Sort
Warp-aggregated scatter reduces atomics and hits 344.6M elements/sec on large benchmarks.
Multiclass Vulnerability Detection (GGNN)
Trained GGNNs on interprocedural program graphs (150K+ funcs); improved macro-F1 0.72 -> 0.84.
Work Experience
Generative AI Research Assistant - [U]Tech
Apr 2025 - Jan 2026Impact-first AI systems work.
- Developed a memory retaining retrieval-augmented generation agent which acts like a patient with predefined medical conditions.
- Built a library-query chatbot reducing front desk workload by 30%.
- Engineered an agent that retrieves NLM images and auto-generates MCQs (~60% professor workload reduction).
- Curated and evaluated 5-10 AI tools monthly; authored adoption reports reviewed by leadership.
Digital Accessibility Assistant - [U]Tech
Jan 2025 - Apr 2025Quality assurance and structured outputs at scale.
- Processed and validated digital content with QA standards to ensure clean, structured outputs.
- Corrected 100+ hours of content to pass accessibility rubrics and metadata accuracy checks.
