Michael L.
Backend & AI Engineer
Experience
Jan 2023 - Oct 2025
2 years 10 monthsLos Angeles, United States
Backend & AI Engineer
Meta
- Designed and implemented the Dreaming Worker pipeline to analyze uploaded user content (audio, PDF, markdown, DOCX) and generate structured “Moments” and “Dreams” via GPT-4-class LLMs.
- Built a multi-tenant content ingestion system with SeaweedFS (S3-compatible), NATS JetStream, and PostgreSQL/TiDB to manage large-scale asynchronous processing.
- Developed file upload and moment creation APIs using FastAPI (Python) integrating LLM agents for automated insight extraction.
- Implemented CLI orchestration commands to run per-tenant or batch analysis, including model selection, lookback windows, and email digest generation.
- Created automated email reporting via templated HTML summaries of generated moments using SMTP server.
- Developed RAG-style pipelines for chunking, embedding, and retrieval of multi-format content (PDF, DOCX, Markdown, audio).
- Developed agentic AI systems that interface with OpenAI, Anthropic, and Claude APIs to process user content and generate structured insights.
- Integrated OpenAI/Claude embeddings using Python and PostgreSQL or TiDB vector indexes to enable high-performance similarity search.
- Integrated Pydantic validation models to enforce structured LLM outputs and ensure schema compliance across the AI insight pipeline.
- Enhanced developer experience with async I/O, structured logging, and streaming agent responses for real-time monitoring of content analysis.
- Built a secure, event-driven file pipeline using SeaweedFS S3 and NATS JetStream, supporting S3/local storage, automated embeddings, and LLM-powered summarization for intelligent indexing.
- Created scheduler and worker systems to automate daily content summarization and insights generation per tenant using NATS JetStream.
- Enhanced CI/CD security with Trivy scans, Pytest automation, and coverage reporting in GitHub Actions.
- Implemented automated unit testing (Pytest), integration testing, and end-to-end UI testing (Selenium), maintaining test quality and preventing regressions across key booking workflows.
Jun 2019 - Jan 2023
3 years 8 monthsLos Angeles, United States
Backend & AI Engineer
Amazon
- Led cross-functional teams leveraging ML, Gen AI, and NLP for financial data analysis and AI-powered investment insights.
- Implemented backtesting and generative AI chat assistant for user education and informed decision-making.
- Led the development of a legal chatbot powered by LLMs, enhanced with agentic AI technology, and deployed on GCP using Kubeflow and BigQuery.
- Demonstrated expertise in cloud-based NLP solutions (Google Cloud, AWS).
- Created interactive dashboards for financial data communication (Tableau).
- Extensive experience implementing cutting-edge research in generative AI (LangChain, ChatGPT, Ollama.ai).
- Scaled distributed backend systems supporting 20,000 daily active users using FastAPI and Node.js, improving request latency and throughput by 20-40% via API refactoring, advanced caching strategies, and database query optimization (MySQL, RocksDB, Resto).
- Developed and maintained enterprise-grade microservices in Python and Django, applying strong multi-threaded programming, data structures, and collections for low-latency, high-availability systems.
- Built and maintained responsive, user-friendly dashboards and client portals using React and Angular, tailored for real-time data visualization and reporting.
- Developed cloud-native applications in Azure and Linux-based environments, optimizing deployments and observability with Prometheus and Grafana.
- Worked in Agile and Scrum environments, leading and contributing to sprint planning, daily stand-ups, retrospectives, and backlog grooming.
- Mentored four junior engineers, led architecture reviews, and established best practices for building scalable, maintainable distributed systems in a low-latency, high-availability environment.
- Designed and developed a crypto arbitrage bot using cutting-edge ML algorithms (Python, TensorFlow, GCP).
- Developed automated trading strategies and real-time decision-making for high-speed execution in volatile markets.
Jan 2016 - Jun 2019
3 years 6 monthsSan Francisco, United States
Backend & AI Engineer
VMware
- Provided services for a director in the creation of various data pipelines and ML engineering.
- Worked on machine learning, NLP, AWS, ElasticSearch/OpenSearch and helped create ETL for Global Hydrogen Index with ElasticSearch.
- Utilized BERT for NLP-based deep learning in voice recognition and integrated agentic AI with Graph Attention Networks, Transformers, and LSTM architectures to enable advanced functionality.
- Proficient in utilizing Alteryx to streamline and enhance data analytics processes.
- Designed and implemented complex workflows that integrated data from various sources, transformed and cleaned data, performed advanced analytics, and generated actionable insights.
- Reduced processing time and improved data accuracy, resulting in more informed decision-making for the organization.
- Participated in the design, development, and deployment of scalable backend services and distributed systems supporting high-traffic e-commerce and financial platforms.
- Built and maintained RESTful APIs and microservices using Python, ensuring high availability and performance for millions of daily users.
- Combined DNNs with SVM, KNN, and tree models using grid search; increased performance by 15%.
- Architected and optimized cloud infrastructure on AWS (EC2, Lambda, S3, RDS, DynamoDB) with infrastructure-as-code tools like Terraform and CloudFormation.
- Developed and maintained frontend applications using React and Vite, delivering seamless user experiences for both internal tools and customer-facing products.
- Implemented CI/CD pipelines with GitHub Actions and AWS CodePipeline, enabling rapid and reliable software delivery.
- Integrated basic AI/ML features (such as recommendation engines or chatbot interfaces) using Amazon SageMaker, OpenAI APIs, and Hugging Face models.
- Monitored and improved application reliability using Prometheus, Grafana, and CloudWatch, reducing downtime and improving incident response metrics.
Jun 2014 - Jan 2016
1 year 8 monthsLos Angeles, United States
Machine Learning Engineer
Microsoft
- Built and maintained scalable backend services in Java and Python, integrating with VMware’s virtualization stack and APIs (vSphere, ESXi, vCenter).
- Developed and enhanced RESTful and gRPC APIs to support automation, orchestration, and third-party integrations.
- Used TensorFlow, Keras, Azure (Docker containers), scikit-learn, and various Python libraries to implement machine learning in a distributed containerized fashion under a Jira Agile methodology.
- Built generative adversarial model for anomaly detection. Utilized R for statistical analysis and data visualization in machine learning projects at Bell Flight, focusing on anomaly detection in helicopter health data.
- Built anomaly model, vanilla artificial neural network, Gaussian mixture model, and autoencoder for anomaly detection.
- Leveraged Alteryx and agentic AI to optimize existing data processes, achieving significant time and resource savings.
- Developed automated, AI-driven workflows that reduced manual data entry by 15 hours per week, boosting team efficiency and minimizing errors.
- Automated hyperparameter testing with itertools (pipeline can test hundreds of models in one go).
Jan 2014 - May 2014
5 monthsChampaign, United States
Full Stack Engineer
Cazoodle
- Built responsive frontends using AngularJS, React, HTML5, CSS3, and JavaScript/TypeScript to deliver high-quality user experiences.
- Developed backend services and RESTful APIs using Python and Node.js, integrating with SQL Server and Azure-based services.
- Implemented Azure cloud solutions including App Services, Azure SQL, Blob Storage, and Azure Active Directory for authentication.
- Collaborated in agile teams, working closely with UX designers, product managers, and QA engineers to deliver high-priority features on schedule.
May 2013 - Aug 2013
4 monthsChampaign, United States
Backend Engineer
NVIDIA
- Wrote scripts in Python and Bash to automate data collection, parsing, and storage.
- Supported integration of backend data pipelines with MySQL and NoSQL databases for large-scale data indexing.
- Wrote and optimized scripts in Python to automate data extraction, transformation, and loading (ETL).
- Assisted in debugging and performance tuning for distributed systems handling high-volume web data.
- Collaborated with senior engineers to improve system reliability and scalability for production workloads.
Summary
10+ years of experience as a Full-Stack & AI/ML Engineer building scalable, cloud-native software and implementing state-of-the-art LLM, GenAI, and NLP systems.
Documented experience working on cross-functional teams and building microservices, CI/CD workflows, and real-time data systems using Python, Java, Kubernetes, and Terraform.
Broad experience integrating RAG architecture, LangChain, Hugging Face, GPT-3.5/4, and optimizing end-user experience.
Interested in building end-to-end AI platforms using Python frameworks, deep learning, and robust DevOps and corporate applications.
Languages
English
NativeEducation
Aug 2008 - Dec 2012
University of Illinois
Bachelor of Computer Science · Computer Science · Champaign, United States
Need a freelancer? Find your match in seconds.
Try FRATCH GPT More actions
Similar Freelancers
Discover other experts with similar qualifications and experience
- English
- Deutsch
2025 © FRATCH.IO GmbH. All rights reserved.