Machine Learning Real-Time

ROBEL LAGER

Phoenix, AZ ***** •************@*****.*** • +1-479-***-**** • linkedin.com/robel-lager-484ab2309

SUMMARY

High-skilled Senior Engineer with more than 10 years of experience in Software Engineering, specializing in Machine

Learning, Blockchain and Web Development based on mathematical approach. Proficient in designing, building,

deploying and maintaining Software Architecture. Seeking to leverage technical expertise and collaborative mindset to

contribute to innovative projects in the AI industry.

PROFESSIONAL EXPERIENCE

COGNEX CORP. Natick, MA

Machine Learning Engineer JAN 2023-July 2024

• Built a ResNet50-based object detection model for real-time defect detection in manufacturing, achieving 92%

precision and 94% accuracy at 30 FPS

• Utilized transfer learning techniques to fine-tune models for industrial image datasets

• Integrated Generative Adversarial Networks (GANs) for image restoration, improving model accuracy by 40%

through sensor noise reduction

• Developed an Auto-Encoder model for compressing industrial images and labels, reducing file sizes by 40%

without compromising visual details

• Achieved 30 FPS processing speed for defect detection while maintaining high accuracy

• Developed several intelligent customer support chatbots and agents utilizing GPT-4, Claude, CrewAI and

LangGraph, achieving 95% positive feedback and 98% successful query handling

• Integrated multimodal LLMs capable of interpreting images, PDFs, and Word documents to analyze research

papers and generate automated summaries

• Innovated zero-knowledge machine learning (zkML) to integrate AI models with blockchain, ensuring data

privacy and secure model updates via Ethereum smart contracts

• Designed advanced feature engineering workflows, including statistical techniques, dimensionality

reduction, and normalization, to optimize model accuracy and scalability

• Engineered robust data pipelines for industrial images and sensor data using AWS S3, GCP BigQuery, and

Apache Kafka for real-time streaming

• Deployed and managed machine learning models using Docker, Kubernetes, and AWS SageMaker for seamless

production environments

• Optimized backend services using Flask, FastAPI, and Spring Boot to build scalable APIs communicating with

AI models

• Designed and deployed DeFi algorithms for market trend prediction, increasing transactional efficiency by

35%

• Engineered DApp integrations to support real-time tracking of tokenized assets on the Ethereum blockchain

using smart contracts

• Developed Web3-based ad attribution models leveraging LLMs, improving campaign performance tracking

by 25%

• Utilized Graph Protocol and Chainlink for decentralized querying and secure data retrieval, enabling scalable

AI applications

• Integrated AI-driven market-making strategies with Uniswap V3 liquidity pools, achieving a 20% increased

ROI

• Applied 3D modeling techniques using Unity and Blender to design virtual representations of industrial

machinery for defect detection and predictive maintenance

• Designed digital twins and visualizations to support training and simulation environments in industrial

operations

• Led the creation of a secure, responsive web application using Angular for real-time defect tracking and

model performance visualization

• Collaborated with research and product teams to integrate Hyperledger Fabric into machine learning

workflows, ensuring secure and transparent AI-driven decisions

BANNER HEALTH. Phoenix, AZ

Senior Software Engineer JULY 2020-Jan 2023

• Led the development of 2D-to-3D pose estimation models using deep learning and CNNs to recognize and

track patient movements in ICUs, enabling real-time monitoring for falls and agitation

• Implemented real-time video analysis of ICU camera footage using transformer-based models to detect

critical events, such as patient movements or respiratory distress, with 50ms latency, achieving 92% accuracy

and 95% recall

• Developed motion recognition systems for monitoring ICU patients' physical activities, detecting anomalies

that indicated distress or other medical conditions

• Built BERT-based NLP models to analyze patient-doctor interaction data, predicting emotional sentiment and

improving patient care and communication

• Applied attention mechanisms in deep learning models to enhance the interpretation of medical data, clinical

notes, and patient records

• Developed speech recognition systems integrated with NLP to enable hands-free interaction with medical

devices and patient data, improving workflow efficiency

• Integrated Generative AI voice assistants to allow healthcare providers to query patient data and clinical

information, enabling faster decision-making and improving workflow efficiency

• Engineered IoT device integration with edge computing solutions for real-time data processing in ICU

monitoring systems, reducing centralized processing requirements and enhancing efficiency

• Designed a distributed microservices architecture to handle large volumes of data from medical devices and

sensor networks, improving scalability, fault tolerance, and system performance

• Optimized parallel processing techniques and distributed computing strategies to handle high-throughput

real-time data streams from medical sensors, enabling faster decision-making in healthcare

• Achieved 89% accuracy, 0.92 AUC-ROC, and 87% sensitivity in predictive models for patient outcome

forecasting in ICU settings

• Integrated MLOps practices using Kubernetes for managing scalable deployments and Docker for

containerizing machine learning models in healthcare applications

• Utilized Azure DevOps for automated pipeline orchestration, ensuring smooth deployment and version

control of machine learning models and AI solutions

• Deployed end-to-end AI models using Docker, Kubernetes, and CI/CD pipelines, ensuring continuous

integration and delivery for healthcare applications

• Ensured compliance with HIPAA for healthcare data management by implementing data encryption (AES256)

and secure data access protocols using OAuth 2.0 and JWT tokens

• Applied data anonymization and pseudonymization techniques to protect patient identities while processing

sensitive medical information for AI applications

• Worked with FHIR standards for secure data exchange between healthcare systems, enabling seamless

integration of medical records across platforms

• Implemented robust IAM systems using Azure Active Directory and AWS IAM to ensure only authorized

personnel accessed sensitive health data

• Conducted regular security risk assessments and developed audit logs and alerting systems for tracking

healthcare data access, ensuring regulatory compliance

• Developed and optimized React-based monitoring applications for real-time tracking of patient health

metrics

• Designed an Angular-based web interface for healthcare providers to monitor ICU data streams and receive

alerts for critical events

• Integrated real-time video analysis systems using React for low-latency visualization of critical events, such

as respiratory distress

• Collaborated with healthcare professionals to monitor AI model performance and identify areas for

optimization

• Reduced response time to critical events by 50% through faster processing of sensor data and enhanced

predictive modeling

• Integrated AI models with healthcare decision support systems to trigger alarms and provide real-time

feedback to medical staff

INDATA LABS. Orlando, FL

Software Engineer NOV 2016-July 2020

• Built a PPO-based reinforcement learning trading bot that dynamically adjusts strategies for high-frequency

trading, achieving a 25% increase in annualized return, 30% reduction in maximum drawdown, and under

2ms execution latency

• Incorporated XGBoost and LSTM models for price prediction and market forecasting, providing actionable

insights to financial analysts

• Applied wavelet analysis for signal processing in time-series data, enhancing stock price prediction accuracy

and reducing noise in financial signals

• Developed feature extraction techniques using Fourier transforms and wavelet transformations to better

capture patterns in volatile market data

• Built a transformer-based NLP model to analyze customer feedback for service improvement, integrating

hypothesis testing and regression analysis to identify key factors influencing customer satisfaction

• Achieved a 15% increase in customer satisfaction scores within 6 months by identifying key pain points and

improving customer service interactions

• Designed distributed microservices and edge computing systems to monitor financial transactions and

predict market trends in real-time

• Leveraged Apache Kafka and Spark Streaming to build scalable real-time data processing pipelines for high-

throughput environments

• Developed robust backend systems for data collection, storage, and processing, using frameworks like Flask,

FastAPI (Python), and Spring Boot (Java)

• Integrated data pipelines with cloud storage platforms such as AWS S3 and GCP BigQuery to support high-

volume financial transactions

• Built secure APIs for interacting with machine learning models and trading systems, implementing OAuth 2.0

for authentication and ensuring data security

• Designed and developed real-time solutions for market trend monitoring and financial anomaly detection,

ensuring reliability in high-frequency trading environments

• Optimized AI systems for scalability, leveraging cutting-edge techniques like transformer models and

reinforcement learning to drive financial insights

EDUCATION

UNIVERSITY OF TORONTO Toronto, On, Canada

Master of Science in Computer Science; Major in machine learning 2010-2016

SKILLS

Programming Languages: Python, C/C++

Packages & Frameworks: PyTorch, TensorFlow, Keras, Scikit-Learn, OpenCV, FastAPI, Flask, OpenAI API

Tools: NI Vision Builder, Matlab, Docker, Kubernetes, Google Cloud, AWS, Microsoft Azure Field:

Computer Vision, Natural Language Processing, Time-Series Forecasting

Contact this candidate