Juspay Genius

Merchant analytics aggregates: metrics, dimensions, optional filters; chart-friendly responses. **Fits well:** Trends, counts, success rate or volume, comparisons across gateways, methods, or time periods. **Domains:** kvorders, kvtxns, kvrefundtxns, kvoffers, mandateexecutionkv, fulfillmentorders. **Filters on categorical fields** (e.g. payment_gateway, bank, payment_method_type): labels people use often differ from stored tokens. Including `field_discovery_results` alongside `filters` helps the service line up `val` entries with merchant-specific distinct values and return validation context when something still does not match the sampled set. The first entry in each ranked inner list from that lookup is usually the right token to filter on. **No categorical string filters** (time range and breakdowns only, or purely numeric thresholds): `field_discovery_results` can stay unset. **Payload shape and UTC rules:** the readonly markdown analytics guide in this connector goes into intervals and JSON shape in more detail.

juspay_query_analytics

Get details for a specific configured alert by its task UID.

Returns a Markdown playbook for investigating a payment-transaction success-rate drop by comparing an analysis window against a baseline period. Use when the user is debugging an SR dip, conversion decline, gateway-performance shift, or rise in failures. The playbook covers dual-period queries, dimensional drilldowns, error-pattern checks, outage cross-references, and a final report outline; each window must be 7 days or less, in ISO-8601 UTC.

Batch distinct-value lookup for analytics filter dimensions: fuzzy string matching against merchant data for a domain and time window (shape-compatible with Genius DimensionLookupRequest). **Inputs:** `domain` matches analytics query domains; `requests` lists filter fields (not metrics or the time-bucket column) with optional string `queries`—one ranked row per query, or empty `queries` to sample up to the first default_limit distincts as one row. **Response:** `results` hold nested lists ranked with the closest matches first. When feeding analytics filters, the first token in each inner list is usually the one to use; later entries help confirm alternatives with the user. Passing the same JSON object through as `field_discovery_results` on the analytics query request keeps validation tied to that ranked list. **Often unnecessary** when there are no categorical string filters, only numeric thresholds, or the field is high-cardinality / free-text (e.g. order IDs, raw error text)—prefer exact literals there.

Live metrics, dimension names, and filter field names for one analytics domain (portal-backed). Use these strings verbatim when building queries; availability and spelling are domain-specific. **Domains:** kvorders, kvtxns, kvrefundtxns, kvoffers, mandateexecutionkv, fulfillmentorders. For request shape (UTC, granularity, filter JSON) see the readonly markdown analytics guide in this connector. Categorical filter values generally match merchant distincts better when a distinct-value lookup is applied for the same window before querying.

Get conflict settings for payment processing

Get detailed info for a configured gateway by MGA ID

Get configuration schema for a specific gateway

Get general merchant settings

Get mandate/recurring payment settings