lebsral

Use when you want conservative, incremental optimization — making small targeted improvements rather than large changes, useful for already-working programs that need fine-tuning. Common scenarios - your program already works well and you want to improve it without breaking what works, conservative optimization that preserves existing quality, fine-tuning a production program incrementally, or whe

Track which optimization experiment was best. Use when you have run multiple optimization passes, need to compare experiments, want to reproduce past results, need to pick the best prompt configuration, track experiment costs, manage optimization artifacts, decide which optimized program to deploy, or justify your choice to stakeholders. Covers experiment logging, comparison, and promotion to prod

Recursive Language Model (dspy.RLM) that explores large contexts via a sandboxed Python REPL -- the LM writes code, queries sub-LMs, and iterates until it produces a final answer. Use when your input is too large for the context window, the model needs to explore data iteratively, you need recursive self-refinement with code execution, or you have research-style tasks requiring programmatic invest

Know when your AI breaks in production. Use when you need to monitor AI quality, track accuracy over time, detect model degradation, set up alerts for AI failures, log predictions, measure production quality, catch when a model provider changes behavior, build an AI monitoring dashboard, or prove your AI is still working for compliance. Also use when you are seeing silent quality drops in producti

Generate synthetic training data when you do not have enough real examples. Use when you are starting from scratch with no data, need a proof of concept fast, have too few examples for optimization, cannot use real customer data for privacy or compliance, need to fill gaps in edge cases, have unbalanced categories, added new categories, or changed your schema. Also used for create training data wi

Find and merge duplicate records across datasets using AI. Use when deduplicating contacts, merging customer records, entity resolution, matching records across systems, finding duplicate tickets, fuzzy matching company names, CRM deduplication, record linkage across databases, matching people across data sources, identifying the same entity in different formats, contact merge, account deduplicati

Use when the mapping from input to output is straightforward and does not need reasoning steps — simple classification, extraction, formatting, or Q&A where minimal latency matters. Common scenarios - simple classification tasks, basic extraction, format conversion, straightforward Q&A, or any task that does not benefit from chain-of-thought reasoning — when you want the fastest possible LM call.

Use when the task benefits from intermediate reasoning before producing an answer — multi-step logic, analysis, math, or complex classification where direct prediction fails. Common scenarios - classification tasks where the model needs to reason about edge cases, math word problems, multi-step analysis, complex question answering, legal or medical reasoning, any task where thinking before answeri

Use when you need to give DSPy agents tool-calling abilities — wrapping Python functions as tools, building tool-using pipelines, or setting up code execution environments. Common scenarios - wrapping a Python function as a tool for DSPy agents, building tool-using pipelines, setting up a calculator or search tool, giving agents access to databases or APIs, or configuring code execution environmen

Use vLLM for high-throughput production serving of self-hosted models with DSPy. Use when you want production LLM serving, tensor parallelism, multi-GPU inference, batch processing, or high-concurrency self-hosted models. Also used for vllm, vLLM, production serving, high throughput LLM, tensor parallelism, self-hosted production, PagedAttention, local production server, GPU serving, batch inferen

Translate text between languages with AI while preserving brand voice and terminology. Use when translating app copy to Spanish, localizing marketing content, multilingual support tickets, i18n with AI, machine translation with brand voice, translating product descriptions, localizing help docs, batch translating i18n JSON files, glossary-enforced translation, AI-powered localization pipeline, tra

Use W&B Weave for DSPy experiment tracking and observability. Use when you want to set up Weave, W&B, wandb, Weights & Biases, experiment dashboard, weave.op, or team collaboration for DSPy. Also used for weave setup, pip install weave, weave.init, wandb project, W&B experiment tracking, weave decorator, weave.op decorator, wandb dashboard, compare optimization runs, team experiment tracking.