Christine Straub

ƒ - Christine Straub #-ï straubchristine § christinestraub € christinemstraub Summary Senior Machine Learning Engineer with 8+ years of expertise in AI/ML, specializing in Generative AI, LLMs, Computer Vision, NLP, and Robotics. Expert in production-grade ML pipelines, multi-agent systems, and cloud architecture (AWS/GCP). Committed to delivering transformative AI solutions that drive measurable business outcomes through innovative engineering approaches. Education University of California - Berkeley Bachelor of Arts in Computer Science University of California - Berkeley Bachelor of Arts in Cognitive Science Berkeley, CA December 2017 Berkeley, CA December 2017 Technical Skills Areas of Expertise: Deep Learning, Generative AI, Large Language Models (LLMs), Computer Vision, MLOps, Natural Language Processing (NLP), Reinforcement Learning, Robotics, Cloud Architecture, Full-Stack Development. Programming Languages: Python, C/C++, JavaScript, TypeScript. Web Frameworks: React, Node.js, Next.js, Vue.js, Django, Flask, FastAPI. AI/ML Frameworks: PyTorch, TensorFlow, Keras, Scikit-learn, FastAI, Hugging Face, MLflow, Metaflow, FiftyOne, Encord, YOLO, OpenCV. NLP & Generative AI: LangChain, Langflow, LlamaIndex, RAG, GPT-4/3.5, Claude, Gemini, LLaMA, DALL·E, Stable Diffusion, NLTK, SpaCy, Transformers, AutoGAN, BigGAN, CycleGAN, and E-GAN GPT. Data Tools: Pandas, NumPy, Matplotlib, PySpark, Metaflow, Airflow, CUDA. Database Systems: SQL (PostgreSQL), NoSQL (MongoDB), Vector Databases (Pinecone, Faiss, Chroma DB), Redis, Elasticsearch. Cloud & Infrastructure: AWS (S3, EC2, Lambda, RDS, SageMaker), GCP (Vertex AI, BigQuery, AutoML), Snowflake, Databricks, Kubernetes, Docker, CI/CD Pipelines, GitHub Actions. Work Experience Senior Machine Learning Engineer November 2024 – Present RIOS Intelligent Machines Palo Alto, CA • Engineered a 5-stage delegated ops pipeline between FiftyOne and Encord, automating dataset filtering, annotation transfer, and reimport—tripling annotation throughput and closing the training loop. • Built “Thoth Frames Retriever” plugin for FiftyOne enabling batch video frame retrieval with multi-camera support, reducing frame access latency and accelerating robotics video processing. • Developed PyTorch-based factory data loaders with sequence-aware batching and timestamp-based ordering, improving GPU utilization while preserving temporal integrity in training pipelines. • Architected scalable ML pipelines using Metaflow and Kubernetes, integrating YOLO-based frame filtering and confidence threshold tuning to reduce anomaly detection latency across robotic video systems. Senior Machine Learning Engineer May 2023 – April 2025 Unstructured IO San Francisco, CA • Led benchmark study of 10+ Vision-Language Models (VLMs) including Claude, GPT-4o, and Gemini, improving table structure recognition by 10% and boosting image-based text extraction by 15% through model selection optimization. • Designed and deployed a scalable multi-agent orchestration system using Pydantic AI and MCP, enabling collaborative LLM task execution and reducing enterprise document processing time • Built an end-to-end document analysis pipeline integrating layout detection, PaddleOCR/Tesseract OCR, and PDF parsing, achieving 30% faster throughput and higher OCR accuracy across enterprise workflows. • Fine-tuned transformer-based OCR models on 11,000+ domain-specific technical PDFs, improving text accuracy by 10% and reducing missing text cases by 13% across mission-critical enterprise systems. Lead Software Engineer (AI/ML) Sapient Logic July 2023 – Feb. 2024 San Diego, CA • • • [DOD Classified] Architected military-grade OCR system processing 5,000+ field documents daily with 97% accuracy, enabling real-time intelligence extraction from captured images for critical Common Operational Picture (COP) updates in contested environments. [DOD Classified] Engineered multilingual translation system that reduced intelligence processing time by 40% through optimized OCR integration, converting French, Arabic, and Chinese documents to English with 92% semantic accuracy for tactical field operations. [DOD Classified] Designed and implemented a mission-critical intelligence requirements management system that decreased collection-to-analysis time by 65%, enabling semi-automated validation of intelligence assets and streamlining distribution to frontline tactical units. Machine Learning Ops | Machine Learning Engineer May 2022 – July 2023 RIOS Intelligent Machines Palo Alto, CA • Architected and deployed AI-powered robotic workcells that increased manufacturing throughput by 25% across enterprise clients, successfully integrating computer vision systems with existing factory automation workflows. • Developed Custom Computer Vision Algorithms that achieved 98% accuracy in part identification and defect detection, enabling real-time quality control for high-volume production environments. • Designed scalable ETL pipelines processing 10TB+ of compressed image data daily, reducing preprocessing time by 40% while maintaining data integrity for training mission-critical machine learning models. • Worked with cross-functional team that reduced robotic system deployment time from 12 weeks to 4 weeks through standardized integration protocols and modular vision system architecture. Senior Software Architect June 2022 – June 2023 Speechlab AI San Francisco, CA • Architected high-throughput API infrastructure for large language models that scaled to handle 12M+ daily requests, reducing latency by 65% while supporting advanced reasoning capabilities across enterprise applications. • Engineered multilingual translation and accent dubbing system processing content in 8 languages, resulting in 40% user engagement increase and enabling seamless localization for global enterprise clients. • Designed and implemented revenue-generating Paywall system with Paddle integration that increased conversion rates by 30% and reduced payment processing errors by 85%. Natural Language Processing Engineer Mar. 2020 – May 2023 PlusOne Communications Salt Lake City, UT • Pioneered BERT-based call analytics system that achieved 94% accuracy in real-time sentiment analysis across 50,000+ daily customer interactions, reducing false positives by 40% while processing multilingual conversations through optimized CPU/GPU deployment architecture. Senior Software Engineer Technical Lead May 2021 – Feb. 2023 Sapient Logic San Diego, CA • Architected and implemented a HIPAA-compliant Electronic Health Record (EHR) system that reduced patient registration time by 25%, streamlined clinical workflows across 4 healthcare facilities, and secured protected health information through custom multi-factor authentication and role-based access controls. Senior Machine Learning Engineer Jan. 2018 – Feb. 2023 Upwork San Francisco, CA • Spearheaded development of 15+ enterprise-grade AI systems integrating computer vision, NLP, and robotics that delivered measurable business outcomes, including a 40% reduction in manufacturing equipment downtime and 94% accuracy in critical NLP prediction tasks. Architected scalable ML pipelines and recommendation engines for 70+ clients across industries, consistently exceeding KPIs while managing full project lifecycles from requirements gathering to production deployment. Senior Data Engineer April 2022 – April 2023 Memetica San Francisco, CA • Designed and implemented an end-to-end data ingestion pipeline that collected sensitive content from multiple fringe platforms (Gab, Truth Social, 4Chan), incorporating text preprocessing logic that enhanced threat detection capabilities for digital investigations. • Developed a comprehensive monitoring ecosystem combining EFK (Elasticsearch, Fluentd, Kibana) stack with Sentry.io integration, enabling real-time alerting and visualization of system performance that improved incident response time from hours to minutes. Chatbot / NLP Engineer Soul Machines Nov. 2021 – June 2022 San Francisco, CA • Engineered advanced conversational AI platform using RASA and Google Dialogflow CX that achieved 98% intent recognition accuracy across 12,000+ daily user interactions, reducing customer support costs by 35% while enabling complex multi-turn dialogues with automated action fulfillment capabilities and continuous learning from user feedback. Google Data Engineer May 2021 – May 2022 Collegis Education Chicago, IL • Architected real-time ETL infrastructure on Google Cloud that processed 250,000+ daily events from external APIs (Phoneburner, Five9, LMS Canvas), reducing data latency by 85% and enabling critical business intelligence capabilities that drove in operational efficiencies through automated decision workflows. Senior Software Engineer Apr. 2021 – Dec. 2021 Inxeption San Francisco, CA • Engineered precision logistics pricing engine that automated FTL, LTL, and domestic shipment cost calculations across 200+ carriers, reducing quote generation time by 87% while increasing pricing accuracy by 22%, directly contributing to additional revenue through optimized load pricing strategies. Machine Learning Engineer Jan. 2021 – May 2021 Sapient Logic San Diego, CA • [DOD Classified] Architected and deployed a mission-critical security analytics platform using advanced NLP algorithms that mapped Tipping Point’s digital vaccines to the MITRE ATT&CK framework with 94% accuracy, reducing false-positive threat identification by 50% while analyzing 10,000+ potential attack vectors to enhance DOD security infrastructure. Software Engineer Sep. 2017 – Apr. 2021 Moody’s Analytics - RMS Silicon Valley, CA • Worked on geospatial analytics platform that integrated location intelligence with financial risk assessment, enabling portfolio managers to visualize $200B+ in assets against natural disaster probability zones and reducing risk exposure by 18% for institutional clients through data-driven investment decisions. Certifications Deep Learning Specialization (DeepLearning.AI) Machine Learning Specialization(DeepLearning.AI and Stanford University) IBM Data Science Specialization(IBM) Machine Learning Engineering for Production (MLOps) Specialization(DeepLearning.AI) Google Data Analytics(Google) Business Intelligence(Google) Coursera, Coursera, Coursera, Coursera, Coursera, Coursera, - Projects Bitcoin LSTM-Based Prediction Engine: Forecast price movements and market trends, LSTM RNN. DocuWiki AI: LLM-powered document platform using BART, DistilBERT, RAG. Customer Satisfaction Predictor: Built on ZenML, achieved 95% accuracy. Conversational Chatbot: GPT-4o + LangChain bot reduced response time by 25%. RAG Pipeline for Banking: LLAMA-3-7B, Fast AI. Automated Supply Chain Pipeline utilizing ML and Deep Learning models: XGBoost and LGBM for future forecasts. Semantic Similarity Analysis for ATT&CK Framework and Digital Vaccine Mapping: MITRE ATT&CK Framework. Medical Image Analysis and Signal Processing: CUDA and cuDNN to optimize deep learning model inference. Forecasting Solution for Energy Log Server Platform: (LSTM), ANN (MultiLayer Perceptrons), and ARIMA models. Urban Environment Analysis via Satellite Imagery: Segmentation (MaskRCNN) to multi-band images. References Available upon request.