Query acceleration
Extract engine, materialised aggregates and result caching keep dashboards fast at peak hours.
naviai Platform · Financial BI
XinBI
XinBI is enterprise-grade intelligent data infrastructure. AI-driven visual exploration, real-time alerts and fine-grained access control — visible, governed and traceable, and built for business at scale.
Connect once · analyse everywhere
How XinBI works
Most BI stacks stitch together a connector, a modelling tool, a dashboard tool and an alerting script. XinBI covers the whole path in one governed product — so the number on the CEO's phone is the same number the analyst built.
100+ connectors for databases, warehouses, data lakes, files and SaaS APIs. Query live or extract, with incremental sync.
Direct queryExtractCDCJoin, union, group and add calculated fields by drag and drop. A governed semantic layer defines every metric exactly once.
Self-service datasetsMetric centreAsk in plain language or drag and drop. 50+ chart types, drill-down, cross-filtering and ad-hoc analysis — no SQL.
Ask XinBIDrill & linkPivotReal-time alerts on thresholds, trends and anomalies, with AI attribution that explains what moved the number.
Anomaly detectionRoot causeDashboards, big screens, mobile, subscriptions and embedded analytics — insight delivered wherever decisions are made.
SubscriptionsEmbedMobileAI-driven visual exploration
Type a business question in everyday language. XinBI maps it to governed metrics, builds the query, picks the right chart and explains what it sees — and shows its working, so anyone can check it.
Try a question
Visible · Governed · Traceable
Speed is useless if nobody trusts the number. XinBI pairs fast exploration with the controls a regulated business needs.
Watch any metric continuously. When it crosses a threshold, breaks trend or behaves unusually, XinBI notifies the owner within minutes — and tells them what drove the change.
Permissions follow the data, not the dashboard. Define them once on the semantic layer and they apply everywhere — dashboards, AI answers, exports, APIs and embedded views.
Click any figure to see where it came from — source tables, transformations, the metric definition and its owner. Every query, change, share and export is recorded, ready for internal audit and regulators.
Built for scale
From one team's dashboard to thousands of users across markets, subsidiaries and currencies — on the same deployment, without re-platforming.
Extract engine, materialised aggregates and result caching keep dashboards fast at peak hours.
Stateless services scale horizontally; scheduled jobs and alerts fail over without missed runs.
Multilingual interface and AI, multi-currency metrics and time-zone-aware schedules across APAC and beyond.
REST API, JavaScript SDK and secure iframe embedding put analytics inside your CRM, portal or app.
Responsive dashboards, a mobile app and bots in Teams, Slack, DingTalk and Lark — ask and get answers on the go.
Command-centre wallboards for risk, operations and collections floors, refreshed in real time.
Full capability map
Feature-complete against the leading self-service BI suites, so switching to XinBI never means giving something up.
For every role
Build datasets, calculations and dashboards by drag and drop, or just ask. No SQL, no waiting on the data team.
Ask follow-up questions on mobile, subscribe to the numbers you own and get an AI briefing every Monday morning.
Define metrics and permissions in one place, see what's used and what breaks, and cut ad-hoc data requests.
A cockpit of the metrics that matter, alerts on what changed and plain-language explanations of why.
Metric packs and dashboard templates from a team that has run consumer and inclusive lending at scale. Start from proven definitions instead of a blank canvas.
Vintage curves, roll rates, DPD buckets and approval quality, with early-warning alerts by cohort and channel.
VintageRoll rateDPD30+PD / LGDFunnel conversion, cost per approved loan and campaign ROI by channel — joined with Lingque CDP audiences.
CACFunnelROILTVContact and promise-to-pay rates, recovery by stage, strategy A/B results and agent performance.
PTPRecoveryCure rateService levels, handling time, complaint trends and AI quality-inspection results across every channel.
SLAAHTCSATComplaintsP&L, net interest margin, cost of risk and budget vs. actual across entities, markets and currencies.
NIMCoRP&LBudgetRegulatory return preparation, concentration limits and conduct metrics — every figure backed by lineage.
ReturnsLimitsConductWhy XinBI
XinBI's AI runs through the naviai AI Gateway: pick any model, mask sensitive fields before every call, and keep a full audit trail.
Learn more → PRIVATEDeploy XinBI fully on-premise, including air-gapped, with private models — no data leaves your network.
Learn more → AI FDEOur engineers move your key reports, metrics and permission models across and train your analysts, one business line at a time.
Learn more →Right. Analysts build datasets, calculations and charts by drag and drop, or simply ask in plain language. Data engineers can still use SQL and Python where they prefer — both work on the same governed metrics.
XinBI answers from your semantic layer instead of guessing at column names. Every answer shows the metrics, filters and query it used; ambiguous questions get a clarifying question rather than a guess, and your team can add examples to improve accuracy over time.
Raw data isn't needed. The model works with metadata and metric definitions while results are computed in your own database. Sensitive fields are masked before any model call, and you can run entirely on private models inside your network.
Yes. XinBI connects to the same databases and warehouses, and our FDE team helps migrate key dashboards, datasets and permission models — usually starting with one business line and going live within weeks.
As SaaS hosted in Singapore, in a dedicated cloud, or fully on-premise including air-gapped. Pricing is per organisation, based on users and deployment model — talk to us for a quote.
Bring one real question from your business. We'll connect a sample of your data and answer it live — no SQL required.