Browse Skills

Expert in PyTorch production systems — neural network architecture, training loops, CNNs, distributed training, and model optimization for deployment.

Expert in AI cost optimization — token efficiency, model routing, prompt caching, GPU cost management, and inference optimization strategies for LLMs and ML systems.

Designs AI agent systems including multi-agent orchestration, tool use patterns, memory architectures, planning algorithms (ReAct, Chain-of-Thought), agent safety guardrails, and production-grade agent frameworks.

Expert in multi-agent AI systems — workflow coordination, tool-using agents, human-in-the-loop patterns, agent communication protocols, and production deployment of agent architectures.

Diagnoses and resolves neural network training failures including gradient issues, loss plateaus, overfitting, and suboptimal hyperparameters. Provides systematic debugging workflows with actionable fixes.

Design, structure, and optimize production-grade system prompts for large language models using proven architectural patterns, research-backed techniques, and adversarial testing methodologies.

Optimizes ML model deployment for production inference including quantization (INT8/INT4), knowledge distillation, ONNX/TensorRT conversion, model pruning, batching strategies, and edge deployment for reducing latency and cost.

Provides expert guidance on time series forecasting including statistical methods (ARIMA, ETS), Prophet, neural forecasting (N-BEATS, Temporal Fusion Transformer), seasonality decomposition, anomaly detection, and confidence interval estimation.

Builds comprehensive responsible AI frameworks including governance structures, model cards, risk assessment processes, fairness metrics, transparency reporting, and regulatory compliance strategies aligned with the EU AI Act and global standards.

Designs recommendation systems covering collaborative filtering, content-based filtering, hybrid approaches, cold start solutions, embedding models, real-time serving, and A/B testing strategies for personalized user experiences.

Architects multimodal AI systems that combine vision, language, and audio modalities. Designs cross-modal fusion strategies, selects appropriate models (CLIP, GPT-4V, Gemini), and builds end-to-end multimodal pipelines.

Designs and optimizes retrieval-augmented generation systems for customer support operations, including knowledge base ingestion, chunking strategies, hybrid retrieval pipelines, answer generation with citation, and continuous quality improvement loops.

Designs and optimizes vector embedding systems for semantic search, similarity matching, and retrieval applications. Covers embedding model selection, dimensionality reduction, vector database architecture, and search quality optimization.

Architect production-grade AI applications with prompt chaining, agent patterns, guardrails, cost optimization, and multimodal capabilities for reliable LLM-powered systems.

Design reliable structured data extraction and generation systems powered by LLMs, with JSON Schema design, provider-specific configuration, validation pipelines, and production error handling.

Design AI agent systems with precisely calibrated human oversight, mapping every action to the right autonomy level — from fully automatic to synchronous approval — using established frameworks for trust calibration, escalation policies, and audit trail architecture.

Design, tune, and evaluate search systems for maximum relevance using BM25, hybrid retrieval, learning-to-rank, and neural re-ranking with rigorous offline and online evaluation frameworks.

Designs automated machine learning pipelines for model selection, hyperparameter optimization, and neural architecture search. Covers AutoML frameworks (Optuna, Ray Tune, Auto-sklearn), search strategies, and production integration.

Design and implement safety, quality, and compliance guardrail systems for LLM-powered applications, covering input validation, output screening, prompt injection defense, PII protection, and regulatory compliance mapping.

Designs comprehensive LLM prompt evaluation frameworks with tailored metric suites, automated eval pipelines, A/B testing methodology, red-teaming protocols, and regression detection systems using industry-standard tools like RAGAS, DeepEval, and promptfoo.

Designs comprehensive evaluation frameworks for LLM and AI systems, combining automated metrics (BLEU, ROUGE, BERTScore), LLM-as-judge patterns (G-Eval), RAG evaluation (RAGAS), standard benchmarks (MMLU, HumanEval), safety evaluations, and A/B testing methodologies into production-ready evaluation pipelines.

Designs privacy-preserving distributed machine learning systems using federated learning. Covers federated averaging, differential privacy, secure aggregation, communication efficiency, and compliance with privacy regulations.

Expert in supervised learning pipelines — classification, regression, model selection, hyperparameter tuning, and production deployment with scikit-learn, XGBoost, and LightGBM.

Conducts comprehensive bias audits and ethical assessments of AI/ML systems, applying fairness metrics, regulatory frameworks (NIST AI RMF, EU AI Act), and structured impact assessments to identify and mitigate algorithmic harms.