**Kunal Kasodekar**
+1 (602) 862-8872
kkasodek@asu.edu
linkedin.com/in/kunalkas/
github.com/gremlin97

### **EDUCATION**
**Arizona State University, AZ**
*Master of Science in Computer Science* | Aug 2022 - May 2024
GPA: 4.00/4.00
*Courses:* Statistical Machine Learning, Algorithms, Data Processing at Scale, Data-Intensive Machine Learning

**Vellore Institute of Technology, Chennai, India**
*Bachelor of Technology in Computer Science and Engineering* | Jul 2015 - Apr 2019
GPA: 8.73/10.00

### **TECHNICAL SKILLS**
- **Programming/Scripting:** Python, C++, SQL, JavaScript, R, Bash
- **ML Frameworks/Tools:** PyTorch, Pandas, Keras, OpenCV, NumPy, TensorFlow, TorchGeo, CUDA, Scikit-Learn, HuggingFace
- **Platforms:** Azure, AWS, Terraform, Docker
- **Data Science:** Hypothesis Testing, A/B Testing, PySpark, Airflow, MLFlow, Data Cleaning & Visualization, Feature Engineering
- **Databases:** PostgreSQL, SQL, Neo4j, RocksDB
- **Other:** Qiskit, Ethereum, Blender, UE4, DVC, QGIS, HuggingFace, GEE, Django, Robot, LangChain, LlamaIndex

### **EXPERIENCE**
**Kerner Lab (Prof. Hannah Kerner), Tempe, AZ**
*Research Volunteer* | Aug 2023 - Present
- Detected and segmented cones on Mars using ensemble segmentation models (UNet, DeepLabv3, FPN, MANet). Improved results with 4-directional overlap, test-time augmentations, and CRF post-processing.
- Developed multimodal segmentation architectures with text-conditioning.
- Created text prompts from research papers using ChromaDB, BGE embeddings, LangChain, and Llama-2.
- Fine-tuned Stable Diffusion (RemoteDiffusion) for remote sensing image generation.
- Generated synthetic datasets for land-use classification, achieving F1=0.56 with TorchGeo.

**W.P. Carey School of Business, Tempe, AZ**
*Research Aide* | Aug 2023 - Jan 2024
- Developed neural inference engines in C++ for gun detection in CCTV streams, deployed on AWS GPU instances.
- Optimized models (FastViT, MobileViT) via quantization, reducing size by 40%.
- Implemented XAI models, reducing false positives by 15%.

**Rocket Central, Detroit, MI**
*Software Engineer Intern* | May 2023 - Aug 2023
- Built a web crawler using Terraform, GCS, and NestJS.
- Created an inpainting demo (Stable Diffusion) for customizing real estate listings.

**Wipro Limited (CTO Office), Bengaluru, India**
*Senior Software Engineer* | Jul 2019 - Jun 2022
- Detected bark-beetle-infested trees using YOLOv3 (mAP=0.507), deployed via Airflow and Azure IoT.
- Developed a simulated forest environment in Blender/Unreal Engine for data generation.
- Implemented LSTM-based wastewater tank forecasting (RMSE=0.34).
- Led an ML/DL team and published tech blogs on quantum computing.

### **MODEL ARCHITECTURE PROJECTS**
- Fully convolutional ViT (65% accuracy on CIFAR-10).
- Custom Fast ResNet (88% accuracy on CIFAR-10).
- Edge-deployable vision model (99.48% accuracy on MNIST).
- Fine-tuned ControlNet for dog image generation.

### **ACADEMIC PROJECTS**
- Multimodal CLIP for satellite images (validation loss=3.34).
- Bias mitigation in COCO dataset (6.5% reduction).
- Transformer-Encoder with relational database (NetsDB).
- AWS facial recognition using Lambda/DynamoDB.

### **PUBLICATIONS & ACCOMPLISHMENTS**
- Published papers on AI budget enforcement (Best Industry Research Paper, SPM Summit 2021).
- Presented at QTML 2021 on quantum feature maps.
- Hosted masterclass on rendering engines for ML data (Wipro TechTonic 2022).
- IBM Certified Associate Developer (Quantum Computation).

no comments

✔ list