3. Data, Science & AI

Guides DNA insertion primer design for site-directed mutagenesis, including verification strategies and procedural guidance for plasmid insertion using Q5 kits.

Guides frame-level video analysis, event detection, and motion pattern processing using OpenCV and computer vision techniques.

Guides recovery of PyTorch models from state dictionaries, fine-tuning specific layers, and saving in TorchScript format for deployment.

Guides accurate document text extraction and parsing using LlamaParse, requiring llama_cloud_services and API key configuration.

Enables cell segmentation using SAM, mask-to-polygon conversion, microscopy data processing, and CLI tool development for biological image analysis.

Processes financial documents via OCR, extracts data, and generates structured summaries (CSV/JSON) while ensuring data safety to prevent loss.

Extracts hidden text from GCODE files by analyzing toolpath geometry and coordinate data, useful for 3D printing, engraving recovery, and CTF challenges.

Designs optimized fusion protein gBlock sequences by integrating multiple protein sources from PDB, plasmids, and fluorescent databases with custom linkers and codon optimization.

Guides accurate token counting in datasets from HuggingFace, ensuring proper schema understanding and avoiding data processing pitfalls.

Guides implementation of Adaptive Rejection Sampling (ARS) for log-concave distributions, covering envelope construction, adaptive updates, and performance considerations.

Guides writing SPARQL queries to retrieve and manipulate academic data from RDF/Turtle datasets, including filtering and aggregation.

Analyzes chess board images to detect piece positions and calculate optimal moves, solving puzzles and providing strategic guidance.

Guides building Caffe from source, preparing CIFAR-10 datasets, and training CNN models with CPU-only deep learning workflows.

Guides reorganization of datasets into shards with constraints on file size and directory contents, optimizing for distributed AI/ML data processing.

Optimizes largest eigenvalue computation for small dense matrices, exceeding numpy/scipy performance using Cython and LAPACK.

Guides training of FastText text classification models with optimized hyperparameters for size-accuracy trade-offs.

Provides guidance for Bayesian MCMC sampling using RStan, including model specification, diagnostics, and hierarchical modeling tasks.

Guides induction proofs in Coq for arithmetic properties like addition commutativity using standard lemmas.

Guides merging heterogeneous data sources (JSON, CSV, Parquet, XML) into unified datasets with conflict resolution for ETL and data consolidation.

Guides recovery of SQLite data from corrupted, encrypted, or inaccessible WAL files, preventing data fabrication and ensuring accurate restoration.

Guides conversion of RStan to PyStan code, including API mapping, hyperparameter translation, and output format handling for Bayesian statistical models.

Optimizes MuJoCo MJCF model configurations for improved simulation performance while preserving numerical accuracy in robotics and physics simulations.

Guides querying ML model performance benchmarks from MTEB and HuggingFace leaderboards, including live data access and avoiding outdated sources.

Guides recovery of data from corrupted or truncated SQLite databases using binary analysis and manual parsing when standard tools fail.