Overview
Prospectis captures inbound B2B leads, qualifies them, removes duplicates, and routes each one to the right partner under a multi-tenant model where every partner sees only their own data. AI scoring grades incoming leads on fit and intent so the strongest opportunities surface first, while quota and credit handling governs how many leads each partner receives and bills for. The result is a faster, cleaner pipeline: partners get well-matched, de-duplicated leads quickly, and the business keeps tight control over distribution and entitlements.
Key Features
Multi-Tenant Isolation
Each partner operates in their own isolated space, enforced at the database level with PostgreSQL Row-Level Security so one tenant can never see or touch another tenant's leads.
AI Lead Scoring
Incoming leads are graded on fit and intent by an AI scoring step, pushing the highest-quality opportunities to the top of each partner's queue.
Duplicate Detection
Matches incoming leads against existing records to suppress duplicates before they reach a partner, protecting both lead quality and credit spend.
Smart Routing
Routes each qualified lead to the best-matched partner based on configurable rules, so the right opportunity reaches the right tenant automatically.
Quota & Credit Handling
Governs how many leads each partner can receive and bill for, with credits provisioned and refunded automatically so entitlements stay accurate.
Edge-Powered Processing
Capture, scoring and routing run on Supabase Edge Functions close to the request, keeping speed-to-lead low even under bursty inbound volume.
Challenges & Solutions
ChallengeHard tenant isolation
A multi-tenant CRM holding competing partners' leads cannot afford even a single cross-tenant data leak.
SolutionSolution Implemented
Enforced isolation in the database itself with Row-Level Security policies, so access is restricted at the data layer rather than relying on application code to remember to filter every query.
ChallengeCatching duplicates that don't match exactly
The same lead often arrives via different forms or sources with small variations, so a strict equality check missed real duplicates.
SolutionSolution Implemented
Applied normalisation and fuzzy matching on the fields that identify a lead, so near-identical submissions are recognised and suppressed before a partner is charged for them.
ChallengeKeeping credits accurate under failure
If a lead failed to deliver after a credit was taken, partners risked being charged for something they never received.
SolutionSolution Implemented
Made credit provisioning idempotent and added automatic refunds on delivery failure, so entitlements always reconcile to what was actually delivered.
Tech Stack
Key Results
- Routed qualified leads to the best-matched partner automatically, improving routing accuracy.
- Suppressed duplicate submissions before they reached partners, protecting lead quality and credit spend.
- Reduced speed-to-lead by processing capture, scoring and routing on edge functions.
- Guaranteed tenant data isolation through database-level Row-Level Security.
Inspired?
Let's discuss how we can build something similar for your business.
Contact Us Today
We design, build, and deploy AI-powered solutions that automate, optimise, and scale your business.