Ishaani Jain

ishaani.jain1@gmail.com Seattle, WA & Irvine, CA U.S. Citizen linkedin.com/in/ishaanijain ishaani.site

Technical Skills

Systems: C++, CUDA, Python, Java — low-level optimization, memory management, performance profiling
ML Infrastructure: PyTorch (C++ extensions), TensorFlow, CUDA kernel development, NCCL, MPI, TensorRT
Hardware: Cache simulation, performance analysis (Nsight Compute, perf), Roofline modeling, memory hierarchy optimization
Cloud & Tools: AWS (Lambda, CDK, DynamoDB, S3, API Gateway, Bedrock, Cognito), Docker, Linux, Git, Next.js, React, Jest

Experience

Undergrad AI/ML Researcher

Feb 2026 – Present

Calit2 · UC Irvine

Built physics-based CV detection algorithm: 100% accuracy, 0 false alarms on sub-pixel fires at 6.8–10.7 miles; deployed for ALERTCalifornia's 1,000+ camera network
Engineered hyperspectral pipeline processing 500-band datacubes (6,788×1,240 px / 2 min) via 1.4 Gbps link; implemented multi-modal fusion (VNIR / SWIR / MWIR / LWIR)
Applied ML/spectroscopy to classify 13+ species (CH₄, CO₂, NO₂, PM2.5) across 40 km — enabled real-time air quality and fire monitoring

Software Development Intern

Feb 2026 – Present

GoFlyy · (3 / 70 accepted, ~4% acceptance rate)

Building CV pipeline for automated garment condition assessment — 90%+ defect classification accuracy
Optimized inference latency 67% (2.1 s → 680 ms) via TensorRT INT8 quantization; building virtual try-on system to reduce return rates

Undergrad Researcher

Mar 2026 – Present

Ramsey Lab · Oregon State University (Hybrid)

Building high-performance tools for the NIH National Center for Advancing Translational Sciences (NCATS) to ingest and query the Babel concept identifier normalization database for the Biomedical Data Translator program
Developed optimized Python/C++ scripts to handle large-scale biomedical data translation

Founding Engineer

Sep 2025 – Present

Stealth Aviation Tech Startup

Architected ML extraction pipeline: 90%+ accuracy on aviation compliance docs using AWS Bedrock; optimized Lambda 65% (8 s → 2.8 s); serving 2 international clients processing 500+ documents
90%+ test coverage across 15K LOC

AI Safety Research Fellow

Jan 2026 – Mar 2026

SPAR · UC Irvine (15% acceptance, 8-week program)

Selected to research reward misspecification, deceptive alignment, and instrumental convergence in vision-language models
Red-teamed GPT-4, Claude 3.5, and Gemini Pro for deceptive capabilities; investigating reward hacking and scalable oversight techniques

Software Engineering Team Lead

Jul 2025 – Sep 2025

C2S Technologies

Led 3-engineer team to win 1st place ($10K seed funding) out of 11 teams; architected full-stack Next.js / Firebase startup matchmaking platform with AI semantic search engine

Projects

Candid — IrvineHacks 2026 Best Overall Winner

Next.js · FastAPI · ChromaDB · scikit-learn · Groq LLaMA 3.3 70B

Won best overall; ingested 52,500+ real legislative chunks across all 50 states into a RAG pipeline powering personalized ballot explanations
Trained linear regression on 2,151 rows of 10-year Census Bureau data (R² = 0.61) to predict household-level financial impact of legislation

High-Performance CUDA Kernels for Matrix Operations

C++ · CUDA · Python · PyTorch

3× throughput via shared memory tiling, warp primitives, and bank-conflict elimination; 40%+ bandwidth reduction through register blocking and coalescing
Reaches 86% of NVIDIA cuBLAS at large matrix sizes; integrated as PyTorch C++ extension; validated via Nsight Compute

Image Classification with CNNs

Python · PyTorch · ONNX

Trained ResNet-18 on CIFAR-10 (60K images, 10 classes) with data augmentation, LR scheduling, and early stopping — achieved 90% test accuracy; deployed with ONNX for inference

Education

University of California, Irvine

Jul 2025 – Jun 2027

B.S. Computer Science & Physics · GPA 4.0 · Junior Standing as Freshman

Coursework: Computer Architecture · Operating Systems · Data Structures & Algorithms · Discrete Mathematics · C++ · Java · Informatics · Multi-variable Calculus · Mechanics · Linear Algebra · Statistics

Clubs: CS: Bits & Bytes · ICSSC · AI@UCI · QC@UCI · Society of Physics Students · AWS Cloud Club (Outreach Officer)