kunal narang

Spearheaded end-to-end deployment of 14+ ML models across GCP, AWS, Azure, and DigitalOcean using Docker, Kubernetes, and MLflow, ensuring fault-tolerant, low-latency inference services. Designed and deployed microservices for CBCT-based ML models (Panoramic Curve Proposal, Anatomy Segmentation) and 3D mesh processing systems (Margin Line Detection), achieving sub-100ms inference latency. Mentored

Engineered ML infrastructure for the Find Intent model in Facebook Search, enabling real-time integrity enforcement based on user intent predictions. Developed precision/recall label generation pipelines and automated dashboards to monitor model performance drift post-launch, identifying a 12% decline in precision and delivering actionable feedback for model retraining. Led critical search interve

Developed sequence tagging and parsing systems using Conditional Random Fields (CRFs) to extract and normalize entities from unstructured records, improving parsing accuracy for downstream analytics. Designed NLP pipelines to compute similarity between email texts using TF-IDF vectors and Word2Vec embeddings, enabling intelligent de-duplication and customer behavior profiling. Conducted fraud dete

Supported development of (5411), an internal ML toolkit enabling users to run classification, regression, image classification, and NLP tasks through a modular interface. Contributed to (Yeti), a custom image recognition dataset and model pipeline for offensive vs. non-offensive image classification, exploring model performance with CNN-based architectures. Performed hyperparameter tuning on vario

Developed and deployed an end-to-end deep learning solution as part of Autodesk’s Digital Finance Transformation initiative, focusing on financial transaction classification using SAP data. Designed and implemented data cleaning, transformation, and feature engineering pipelines in Python to prepare transactional records for model training. Deployed the trained model to SAP Leonardo Machine Learni

Pursued a Master's Degree in Computer Science, specializing in advanced topics. The program enhanced skills in various areas of computer science.

Completed a Bachelor's Degree in Computer Science, gaining foundational knowledge in the field. The curriculum covered core computer science principles and programming concepts.