Leonid Ganeline
Verified Expert in Engineering
Software Development Lifecycle (SDLC) Developer
Vancouver, BC, Canada
Toptal member since September 18, 2024
Leonid is a machine learning and data science engineer proficient in data exploration, experimentation, model training, and fine-tuning using Python, SQL, and Cloud ML. With experience in natural language processing (NLP), anomaly detection, and expertise in building ML teams, Leonid is ready for his next challenge.
Portfolio
Experience
- Software Development Lifecycle (SDLC) - 15 years
- Machine Learning - 10 years
- System Architecture Design - 10 years
- Natural Language Processing (NLP) - 9 years
- Data Science - 8 years
- Anomaly Detection - 3 years
- Named-entity Recognition (NER) - 3 years
- LangChain - 2 years
Work Experience
AI/ML Engineer
Renes Trading LLC
- Created the 3D investor classification that allows flexible calculations of the risk score and ROI. See a blog post (https://www.serava.com/blog/real-estate-investor-dna-framework).
- Developed the AI investor adviser, which generates investor reports that help with property investment decisions.
- Developed and trained the price trend model for properties.
Senior Machine Learning Engineer
Stealth Startup
- Created a chat based on retrieval-augmented generation (RAG) using private and public data in different formats.
- Productized this chat with the Chroma vector store and open-source large language models (LLMs).
- Performed excessive evaluation of synthetic data generated by LLMs.
Senior Machine Learning Engineer
Tigera
- Created an anomaly detection model framework for the Calico Enterprise and Calico Cloud products. It included productizing ML models into the Calico Kubernetes clusters.
- Developed classification models based on the CatBoost and tokenizers with novel data preprocessing.
- Built time-series models based on the GluonTS neural networks, Isolation Forest, and local outlier factor (LOF) and ensemble clustering models.
Senior Machine Learning Engineer
SkyHive
- Engaged as the first data scientist at SkyHive. Initiated data science and machine learning projects and created and owned the entire machine learning technology stack, from envisioning to production.
- Developed production services and applications. Utilized word2vec, fastText, and embeddings from language models (ELMo) for classification and text similarity. Established workflows for data labeling, model evaluations, and regression testing.
- Performed labeling and the evaluation of training datasets with Amazon Mechanical Turk (MTurk).
- Implemented REST services and deployed them with Azure DevOps pipelines and Kubernetes in Azure, Google Cloud, and AWS. Reviewed code and hired for the ML team.
BizTalk Developer
Visiphor Corporation (former Sunaptic Solutions)
- Developed complex XML transformations on Extensible Stylesheet Language Transformations (XSLT) and XML Schema Definition (XSD).
- Built SQL queries and stored procedures that are used in the BizTalk adapters.
- Designed message orchestrations to transfer messages between systems.
Project History
LangChain Project Contributor
https://github.com/langchain-ai/langchainLinks:
https://github.com/langchain-ai/langchain/graphs/contributors?all=1
https://python.langchain.com/docs/people/
Anonymizer
https://github.com/leo-gan/anonymizerSee documentation: leo-gan.github.io/anonymizer/101/why-anonymization
Multi-language Serialization Benchmarks and Course
https://leo-gan.github.io/GLD.SerializerBenchmark/theory/101/Processing Legal Case Bundles
The project's goal was to parse these huge PDF files, extract, and analyze all case information.
Tools:
• Hugging Face Gradio - for the chat UI
• LangGraph - for the agent interactions
• Google Gemini models - for parsing, extraction, analysis
Legal Document Processing
Code Correcting Agent
AI Agent to Develop Software
AI-based Real Estate Platform
http://serava.comEducation
Master's Degree in Electronic Engineering (Signal Processing)
Samara State Aerospace University - Samara, Russia
Certifications
Reinforcement Fine-Tuning LLMs With GRPO
DeepLearning.AI
Post-training of LLMs
DeepLearning.AI
Fundamentals of MCP
Hugging Face
Fundamentals of Agents
Hugging Face
Build Apps with Windsurf’s AI Coding Agents
DeepLearning.AI
Practical Multi AI Agents and Advanced Use Cases with crewAI
DeepLearning.AI
AI Agentic Design Patterns with AutoGen
DeepLearning.ai
Building AI Applications With Haystack
DeepLearing.AI
Long-Term Agentic Memory With LangGraph
DeepLearning.AI
Functions, Tools and Agents with LangChain
DeepLearning.AI
Quantization Fundamentals with Hugging Face
Deeplearning.ai
Retrieval Optimization: Tokenization to Vector Quantization
Deeplearning.ai
Vector Databases: from Embeddings to Applications
DeepLearning.AI
LangChain for LLM Application Development
DeepLearning.AI
LangChain Chat with Your Data
DeepLearning.AI
Large Language Models with Semantic Search
DeepLearning.AI
How Diffusion Models Work
DeepLearning.AI
Data Manipulation at Scale: Systems and Algorithms
University of Washington
Neural Networks for Machine Learning
University of Toronto
Machine Learning
Stanford University
Skills
Libraries/APIs
NumPy, Scikit-learn, Pandas, SpaCy, Natural Language Toolkit (NLTK), Hugging Face Transformers, OpenAI API, XGBoost, Pydantic, REST APIs, PyTorch, Keras, Gradio
Tools
PyCharm, Jira, GitHub, Named-entity Recognition (NER), ChatGPT, Make (formerly Integromat), AI Prompts, Git, Slack, Amazon SageMaker, Windsurf, Haystack, Claude Code
Languages
Python, SQL, C, XSLT, C#, XML, XSD, Rust, JavaScript, Java, Swift, Go, PHP, Kotlin, Zig
Platforms
Azure, Linux, Amazon Web Services (AWS), Google Cloud Platform (GCP), Kubernetes, Docker, AWS Lambda, CrewAI, Ollama
Storage
JSON, Data Pipelines, Elasticsearch, MongoDB, Google Cloud
Industry Expertise
System Development Life Cycle (SDLC), Healthcare
Frameworks
LangGraph, Agentic Frameworks, LightGBM, .NET, MXNet, LlamaIndex, AutoGen, Django
Paradigms
Anomaly Detection, Azure DevOps, MapReduce, HL7 FHIR Standard, Model Context Protocol (MCP)
Other
BizTalk Server, Natural Language Processing (NLP), Data Science, Large Language Models (LLMs), LangChain, Retrieval-augmented Generation (RAG), Artificial Intelligence (AI), Machine Learning, API Integration, Deep Learning, Natural Language Understanding (NLU), AI-assisted Development, Software Development Lifecycle (SDLC), AI Tools, Back-end, FFT, Algorithms, Radio, Software Development, Cloud Computing, Embedding Models, Open-source LLMs, AI Agents, Signal Processing, Generative Artificial Intelligence (GenAI), Supervised Learning, Prompt Engineering, OpenAI, Clustering, Clustering Algorithms, Model Tuning, AI Model Training, OpenAI GPT-4 API, Document Processing, Gradient Boosting, Data Processing, Minimum Viable Product (MVP), Proof of Concept (POC), Hugging Face, ChatGPT Prompts, ChatGPT API, Data Preprocessing, Technical Leadership, Architecture, Feature Engineering, Data Classification, Data Analytics, Vector Databases, Document Parsing, Text Classification, Unsupervised Learning, System Architecture Design, Leadership, Data Mapping, Data Engineering, APIs, Logistic Regression, Cloud, Software Architecture, Time Series Data, Model Validation, Quantization, AI Chatbots, Vector Search, Context Engineering, AI Architecture, Model Evaluation, RAG Architecture, AI Integration, Web Scraping, Taxonomy, AI Enablement, Chatbots, Chatbot Conversation Design, Real Estate, Statistics, Linear Algebra, GluonTS, Poetry, Information Retrieval, Software Design, Vector Stores, Ruff, Deep Neural Networks (DNNs), Neural Networks, Evaluation, Weaviate, Digital Signal Processing, Electronics, Mathematics, Mathematical Analysis, Reinforcement Learning, Pinecone, Computer Vision, FastAPI, Transformers, Small Language Models (SLMs), Agentic AI, Generative Systems, Data Scraping, Mathematical Statistics, smolagents, Json Patch, Ontologies, PDF, ChromaDB, Qdrant, Tokenization, GRPO, Fine-tuning, Reinforcement Learning from Human Feedback (RLHF), Google Gemini, Google GenAI, Model Monitoring, Model Deployment, Cursor AI, Data Anonymization, Optical Character Recognition (OCR), AI Pipeline, Predictive Analytics, Recommendation Systems, Modeling, Gemini, Anthropic, LLM Integration, C Plus Plus, Analysis, Mojo.js
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring