Train, evaluate, and improve AI systems that understand cryptocurrency, blockchain, and DeFi. Remote, full-time or contract, mid-level to senior.
About the role
We are looking for an experienced AI Model Training Specialist with strong knowledge of cryptocurrency, blockchain technology, and decentralized finance. You will help develop, train, evaluate, and improve AI systems capable of understanding complex crypto-related information.
This is a fully remote position for someone who combines machine-learning expertise with a practical understanding of blockchain ecosystems, market structure, security, and digital assets.
Responsibilities
- Build and curate high-quality datasets for training and fine-tuning AI models.
- Design supervised fine-tuning, preference optimization, and retrieval-augmented generation workflows.
- Develop taxonomies and annotation guidelines for cryptocurrency and blockchain data.
- Train models to interpret technical documentation, governance proposals, smart contracts, market data, and on-chain activity.
- Create difficult evaluation datasets covering factual accuracy, reasoning, security, and hallucination resistance.
- Design adversarial tests involving scams, prompt injection, market manipulation, wallet-draining attacks, and misleading financial claims.
- Implement automated and human-in-the-loop evaluation pipelines.
- Measure model quality using precision, recall, F1, calibration, ranking metrics, and task-specific benchmarks.
- Improve model performance through prompt engineering, data augmentation, fine-tuning, and systematic error analysis.
- Build RAG pipelines using vector databases, embeddings, rerankers, and trusted data sources.
- Work with structured and unstructured information from block explorers, APIs, white papers, protocol documentation, and governance forums.
- Detect dataset contamination, label leakage, class imbalance, duplication, and temporal inconsistencies.
- Collaborate with engineers, researchers, analysts, and subject-matter experts.
- Document experiments, dataset provenance, model limitations, and reproducibility procedures.
Required qualifications
- Professional experience training, fine-tuning, evaluating, or deploying machine-learning models.
- Strong Python skills and experience with PyTorch, Hugging Face, or comparable frameworks.
- Practical experience with LLMs, transformers, embeddings, tokenization, and context-window management.
- Understanding of supervised fine-tuning, LoRA/QLoRA, preference optimization, and model evaluation.
- Experience processing large datasets using SQL, pandas, Spark, or similar technologies.
- Familiarity with experiment tracking, model versioning, and reproducible ML pipelines.
- Strong understanding of cryptocurrency fundamentals, including:
- Bitcoin and Ethereum
- Wallets, keys, signatures, and transaction lifecycles
- Smart contracts and token standards
- DeFi, liquidity pools, lending, staking, and bridges
- Layer 1 and Layer 2 networks
- DAOs and governance
- Stablecoins, exchanges, custody, and market structure
- Ability to distinguish technical facts from promotional claims and financial speculation.
- Strong written communication and the ability to explain complex concepts clearly.
- Ability to work independently in an asynchronous remote environment.
Preferred qualifications
- Experience working with on-chain analytics or blockchain data-indexing platforms.
- Familiarity with Solidity, EVM execution, smart-contract auditing, or Web3 libraries.
- Experience with distributed training, quantization, inference optimization, or GPU infrastructure.
- Knowledge of MLOps tools, Docker, Kubernetes, CI/CD, and cloud platforms.
- Experience building agentic AI systems and tool-calling workflows.
- Familiarity with vector databases such as pgvector, Pinecone, Weaviate, Milvus, or Qdrant.
- Understanding of cryptographic primitives, zero-knowledge proofs, consensus mechanisms, and cross-chain communication.
- Experience addressing model security issues such as prompt injection, data poisoning, jailbreaks, and sensitive-information leakage.
- Published research, open-source contributions, or demonstrable AI/crypto projects.
Example projects
You may work on projects such as:
- Training an AI assistant to analyze blockchain protocols and technical documentation.
- Creating a benchmark for evaluating cryptocurrency reasoning and factual accuracy.
- Developing a RAG system backed by verified, time-sensitive crypto information.
- Building classifiers for scams, phishing attempts, malicious tokens, and suspicious transactions.
- Teaching models to explain smart-contract behavior while clearly communicating uncertainty.
- Evaluating whether model-generated financial information is current, sourced, and appropriately qualified.
What we value
- Intellectual honesty and careful handling of uncertainty
- Security-first thinking
- Strong technical judgment
- High-quality documentation
- Responsible treatment of financial information
- Curiosity about emerging AI and blockchain technologies
- Clear, proactive remote communication
How to apply
Please submit:
- Your résumé or LinkedIn profile
- A brief introduction
- Links to relevant GitHub repositories, research, or portfolio projects
- A description of an AI model you have trained or evaluated
- An example of your experience with cryptocurrency or blockchain technology
- Your availability, time zone, and compensation expectations
