Operant Labs
From single-tenant startup to enterprise-ready platform.
A multi-tenancy, SSO, and usage-metering rebuild that unblocked eleven stalled enterprise deals in the first two quarters after launch.
- Year
- 2025
- Duration
- 8 months
- Team
- 10 people
- Sector
- SaaS & Technology
0
Enterprise deals unblocked
Closed in two quarters post-launch
0%
Security review time
Reduction in average days to clear
0.00%
Billing accuracy
Metered usage reconciled to invoice
0%
Tenant isolation
Queries enforced at the data layer
Operant Labs
A B2B analytics platform serving operations teams at mid-market and enterprise companies, growing from 40 to 380 customer accounts during the engagement.
To give operations teams the analytical tooling that engineering teams take for granted.
Logo concept
An operator bracket enclosing a rising signal — mathematics as the product's foundation.
#4C1D95#8B5CF6Founded
2020
Headquarters
Denver, Colorado
Employees
95
Sector
SaaS & Technology
Every surface we shipped
11 designed screens across 2 deliverables, rendered live rather than captured as static images.
Product Marketing Site
Positioning and self-serve trial entry.
Workspace
Tenant-scoped analytics with saved views.
Cohort Explorer
Retention and behavior analysis across segments.
Orders
High-density queue with bulk actions and saved views.
Data Sources
Connector catalog with per-tenant sync status.
Usage & Billing
Metered consumption reconciled to invoice.
Organization Settings
SSO, SCIM provisioning, and audit configuration.
Mobile application
Sign In
Create Account
Notifications
Profile
Across every form factor
The same design system, rendered at each breakpoint it has to survive.
The challenge
Operant had product-market fit and a growing pipeline it could not close. Every enterprise deal stalled at security review: no SSO, no SCIM provisioning, no audit log, and a database schema where tenant separation was enforced in application code rather than at the data layer — a fact that a single competent security reviewer found immediately. Sales had lost seven deals in twelve months for reasons that were entirely architectural.
Research
- Post-mortem interviews on all seven lost enterprise deals
- Review of 23 completed security questionnaires to identify recurring blockers
- Schema audit identifying every query path lacking tenant scoping
- Benchmarking against SOC 2 Type II and ISO 27001 control requirements
The solution
Tenant isolation moved from application code into PostgreSQL row-level security, so a query that forgets its tenant scope returns nothing rather than everything — a failure mode that fails safe. SSO and SCIM shipped as configurable product features rather than per-customer integrations. Usage metering was rebuilt as an event pipeline with reconciliation guarantees, replacing an aggregation job that had been quietly under-billing by a low single-digit percentage.
UX decisions
Self-serve SSO configuration with a test-before-enforce step
Every competitor required a support ticket. Letting an IT admin configure and verify SSO alone removed days from the onboarding path.
Audit log presented as a product feature, not a compliance artifact
Customers were buying it for security review but using it for internal investigation. Designing for the second use made the first trivially satisfiable.
Usage dashboards visible to customers, not just billing
Surprise invoices were the top support driver on the metered plan. Continuous visibility eliminated the category.
Tenant switcher deliberately prominent for multi-org users
Consultants working across client tenants had been signing in and out repeatedly. Acting in the wrong tenant is the worst possible error, so the current tenant is always visible.
Features
- Row-level tenant isolation enforced in PostgreSQL
- SAML and OIDC SSO with self-serve configuration
- SCIM 2.0 user and group provisioning
- Immutable audit log with export and retention policies
- Usage metering reconciled to invoice at the cent
- Per-tenant performance guarantees with query governance
- Customer-facing usage and cost dashboards
Technology
Architecture
A GraphQL gateway resolves tenant context from the session and sets the PostgreSQL session variable that drives row-level security policies — application code cannot bypass it. Metering events flow through an append-only pipeline into a daily reconciliation job that reconciles to the billing system, alerting on any variance above one cent. Redis caches tenant configuration with explicit invalidation on change. Terraform manages the AWS estate; per-tenant query governance sits at the gateway, so a single tenant's expensive query cannot degrade another's experience.
Results
Eleven previously-stalled enterprise deals closed in the two quarters following launch. Average security review time fell 73%, from 41 days to 11. The metering rebuild surfaced and corrected historical under-billing, adding measurable recurring revenue that had been silently leaking. Operant achieved SOC 2 Type II certification four months after the compliance work shipped.
Lessons learned
- 01Enforcing isolation at the data layer rather than in application code was worth the migration cost several times over — it converted a class of catastrophic bug into an impossible one.
- 02The lost-deal post-mortems were more valuable than any requirements document. Sales knew exactly what was blocking them; nobody had systematically asked.
- 03Building audit logging as a usable product feature rather than a compliance checkbox meant it got used, which in turn made it battle-tested before any auditor examined it.
How the engagement ran
8 months across 4 phases with a team of 10.
Enterprise Readiness Audit
4 weeksMapped every blocker cited in lost enterprise deals against the current architecture.
Gap analysisDeal-loss taxonomyPrioritized roadmapMulti-Tenancy Migration
14 weeksIntroduced tenant isolation at the data layer with zero-downtime backfill.
Isolation modelMigration toolingRow-level policiesIdentity & Compliance
10 weeksShipped SSO, SCIM, and audit logging as first-class product features.
SSO integrationSCIM provisioningAudit logMetering & Billing
8 weeksBuilt the usage pipeline with reconciliation guarantees to the cent.
Metering pipelineBilling reconciliationUsage dashboards
Have a problem shaped like this one?
We start every engagement with a paid discovery sprint. You get an architecture assessment and a delivery plan — whether or not you continue with us.
Related engagements
About this case study: Operant Labs is a fictional client. This engagement, its metrics, and its quotes are illustrative work product created to demonstrate our delivery approach, architecture reasoning, and design process. They do not describe a real customer.

