[ THE PROBLEM ]

A build-vs-buy decision at scale

PAR’s internal prototype proved the demand — but productionizing it across 20+ use cases and hundreds of restaurant brands was a different problem entirely.

PAR’s engineering team built an internal analytics prototype that handled foundational use cases and confirmed strong demand from restaurant operators. But the scope kept growing — 20+ use cases spanning multiple platforms, each requiring deeper cross-system reasoning. Scaling the prototype to production quality would have meant dedicating engineering capacity that was needed on PAR’s core product roadmap.

The data environment reflected PAR’s growth through strategic acquisitions — hardware, payments, loyalty, ordering, and engagement cloud platforms each with their own mature data infrastructure. Delivering unified analytics meant reasoning across all of these systems simultaneously, a specialized problem that would require specialized tooling.

The internal prototype worked well for single-platform queries, but the highest-value questions — correlating order trends with loyalty engagement, tying menu performance to payment data — required cross-platform reasoning that’s a fundamentally different engineering challenge. PAR’s leadership saw the tradeoff clearly: continue investing engineering cycles into analytics infrastructure, or partner with a team that had already solved that problem.

[ THE SOLUTION ]

AI-native analytics embedded where customers already work

TextQL built directly into PAR’s existing dashboard experience, turning their analytics prototype into a production-grade product feature.

TextQL’s Ana gave restaurant brands a white-labeled, embedded analytics layer that works across PAR’s multi-platform data environment: hardware, payments, loyalty, ordering, and engagement cloud platforms. PAR serves hundreds of restaurant brands from a shared data infrastructure, so Ana had to answer the right questions for the right people, with no data bleeding across brand boundaries.

Rather than forcing customers to learn new tools, navigate disconnected databases, or stand up separate data environments per customer, Ana unifies PAR’s varied data structures behind a single natural-language interface, with access automatically scoped to each operator’s context.

Secure by design. Scalable by default.

Most embedded analytics deployments solve the multi-tenant problem the hard way: separate databases, separate deployments, separate maintenance burdens for every customer. PAR couldn’t scale that way across hundreds of restaurant brands and 140,000 locations.

TextQL’s Role-Level Security changes the architecture. PAR’s data team uses Client Filter Rules in the Ontology to define exactly which fields, metrics, and data each user role can see. Those rules live in one place; Ana enforces them on every question, every generated report, every live dashboard, automatically.

A regional manager at one restaurant brand asks Ana about their top-performing menu items and gets an answer scoped to their data, no configuration required. Another operator at a competing brand asks the same question and gets their answer too, from their data only. Same deployment, completely isolated outputs.

The business impact runs in both directions. PAR’s customers get analytics they can actually trust, with no risk of accidentally surfacing a competitor’s order volumes or loyalty metrics. And PAR gets an analytics layer that scales without multiplying infrastructure costs. One deployment serves the entire customer base. New brands onboard into the same environment, inherit the access rules automatically, and start querying their data within the hour.

RLS turns a multi-tenant data problem into a product feature. PAR can open the full analytics surface to every operator because the access rules are already enforced at query depth.

• Embedded in the existing dashboard

  Ana lives inside PAR Ordering’s dashboard. No new tool to learn, no separate login. Restaurant operators get analytics where they already work.

• Cross-platform data unification

  Ana reasons across PAR’s diverse data platforms and structures, delivering unified answers regardless of which system the data lives in.

• Role-Level Security at query depth

  Client Filter Rules in the Ontology enforce data boundaries on every interaction. One deployment, hundreds of brands, zero data bleed.

• White-labeled for PAR’s customers

  The experience is branded as PAR’s own. Restaurant operators interact with analytics that feel native to the platform they already trust.

• 60 minutes from onboarding to insights

  New restaurant brands go from first login to their first analytics query in under an hour. Access rules apply automatically. There’s nothing to configure.

The setup was a connector flow. PAR’s data team pointed Ana at their existing data environment, defined the access rules in the Ontology once, and Ana was in production in days. No data left PAR’s environment. The existing tools still worked the day Ana went live. They just started answering questions — and only the right ones.

[ THE RESULTS ]

Engineering hours back on the product roadmap

With Ana embedded natively in the PAR dashboard, restaurant brands operating tens of thousands of locations and processing billions of monthly transactions can interrogate their data directly, without an analyst queue, tickets, or wait times.

The engineering team that had been maintaining the analytics prototype shifted back to core product development. The analytics function became a product feature rather than an internal infrastructure project.

The expansion from the initial deployment into broader use cases happened organically. Once restaurant operators saw what Ana could do with ordering data, they started asking about loyalty, payments, and engagement metrics. The 20+ use cases weren’t planned — they emerged from operators discovering what questions were suddenly possible to ask.

PAR made the shift from treating embedded analytics as an internal engineering project to treating it as a product feature powered by a specialist partner. What would have taken years to productionize in-house was live in days.