What we changed, and what it changed for them.
AI-assisted recruitment workflow inside InterPanely
Hiring teams screening at volume · built and operated by Arrayscript
Recruiters spent most of the day on manual screening and scheduling. Every panel used a different rubric, so candidate comparisons were opinions rather than data.
A workflow that summarises each application against the role's own criteria, cites the CV passage behind every claim, proposes panel slots, and routes anything low-confidence to a recruiter instead of guessing.
We build AI that works.
Grounded in your own data, measured against your own cases, and answerable to a person who signs off.
Stack: Angular · .NET · Postgres · hosted LLM in-region. Evaluation set of 400 historical applications gated every release.
From spreadsheets to one operational system
Orders, approvals and status lived in three spreadsheets and a WhatsApp group.
A single web application with roles, approvals and an audit trail, deployed in stages so work never stopped.
Production application, faster workflows, one source of truth
Quoting that can be traced back
Quotes were verbal or on paper; nobody could reconstruct what a customer had been promised.
Versioned quotation workflow with approval steps and a printable record, built on Quotation Plus.
Same-day quotes, fully auditable history
Calculation accuracy at mobile speed
A consumer tool where a wrong number is worse than a slow one, on mostly low-end devices.
Calculation engine covered by tests, rendered client-side with a strict performance budget.
Sub-second results, Core Web Vitals in the green
Want something like these, but yours?
Start a Project