From research to working systems — prompt builders, library and workflows for RAG, Agents, Multimodal and Text-to-SQL
Prepare RAG, agents, multimodal systems and Text-to-SQL with design and evaluation templates. Build a prompt, run it in your preferred AI tool, then bring the result into a workflow.
Design ingestion, retrieval, reranking, grounded answers and stage-specific evaluation.
Compare workflows with agents; define tool contracts, stopping rules and recovery.
Build held-out cases, rubrics, human calibration and regression checks.
Separate observations from inferences across text, images, audio and video.
Design schema-grounded queries with read-only constraints and validation.
Define field-level evidence, missing-value policies and JSON validation.
Design observability, caching, fallbacks, releases and cost budgets.
Improve prompts from failures, then compare against a fixed evaluation set.
Generate a prompt to find a research breakthrough from multiple angles.
Generate an audit prompt to check code, experiment, and paper quality.
Generate a multi-persona paper review prompt.
Generate a research direction and literature survey prompt.
Generate a systematic experiment design prompt.
Generate a LaTeX paper writing prompt matching venue style.
Generate a venue/timing submission strategy prompt.
Generate a rebuttal strategy prompt for reviewer comments.
Generate a systematic literature review / survey prompt.
Generate a research grant proposal drafting prompt.
Generate a dataset/benchmark design prompt.
Generate a presentation structure/storyline prompt.
Generate a personalized concept-explanation tutor prompt.
Generate a prompt to implement an algorithm, pipeline, or utility.
Generate a root-cause diagnosis prompt from errors/logs.
Generate a hyperparameter search strategy prompt.
Generate a conference poster design prompt.
Generate a thesis chapter/outline prompt.
Generate a model card or dataset datasheet prompt.
Design a data-cleaning pipeline (missing/outliers/duplicates/types).
