"Swiggy released Hermes V3, a GenAI-powered text-to-SQL assistant that enables employees to query data in plain English. Hermes operates within Slack, combining vector retrieval, session memory, agentic orchestration, and an explanation layer to generate accurate SQL queries from natural language inputs. Swiggy, an Indian online food ordering and delivery company, initially launched Hermes as a lightweight interface that allowed employees to ask simple questions and receive corresponding SQL queries executed against internal data stores."
"In its third iteration, Hermes introduces a vector-based prompt retrieval system built on historical SQL executed in Snowflake. Still challenged by most production queries lacking descriptive metadata, the team utilized large-context language models to convert SQL queries into natural-language explanations, effectively reconstructing the missing query intent. These generated prompts are indexed using vector similarity and injected as few-shot examples, allowing Hermes to ground new requests in prior analytical patterns and significantly improve SQL generation accuracy."
Hermes V3 is a GenAI-powered text-to-SQL assistant that enables employees to query data in plain English within Slack. The system combines vector retrieval, session memory, agentic orchestration, and an explanation layer to generate SQL from natural language. Early versions produced inconsistent results, lacked conversational context, and could not validate derived metrics. The rebuilt architecture uses few-shot learning, metadata retrieval, and structured LLM workflows. Historical SQL queries executed in Snowflake are converted into natural-language explanations and indexed via vector similarity to serve as few-shot examples. Grounding requests in prior analytical patterns raises SQL generation accuracy and supports multi-turn interactions.
Read at InfoQ
Unable to calculate read time
Collection
[
|
...
]