The client
People Up, a recruitment firm working candidate placements across European markets and Canada. Their recruiters source talent at scale, research candidates and companies before outreach, run pre-screening, and manage email campaigns across the pipeline.
The brand: https://peopleup.llc
The problem
The recruitment workflow had fractured across too many tools.
Every recruiter was running:
- One platform for scraping candidates
- A different one for company prospecting
- A third for AI research and enrichment
- A CRM that didn't integrate with the above
- An email client that didn't share state with the CRM
- A mock-interview service that lived in its own silo
This is the industry pattern, not a People Up outlier. The average company today runs 106 SaaS applications (BetterCloud State of SaaSOps 2024), and within that sprawl recruiting teams alone typically rely on five to ten distinct systems in their daily workflow.
Five subscriptions, five data silos, five logins. Every candidate touched four or five tools before a recruiter ever made the call. Context got lost at every transition. Decisions got made on incomplete information.
People Up didn't need another tool. They needed one workstation.
Our approach
We built a desktop-grade recruitment OS where every step of the candidate journey runs in one application, and AI is built into every step instead of bolted on after.
AI is no longer optional in talent acquisition. 62% of employers now use AI somewhere in their hiring process, up from 40% in 2020 (Aptitude Research 2025). The same research finds only 6% of companies have automated more than three quarters of their workflow, and recruiters remain wary of AI without transparent governance. People Up's stack puts AI at every step but keeps the recruiter in control of every closing conversation.
Lead Sourcing. Configurable scraping profiles pull candidates and companies from across professional networks, business directories, and geographic searches. Every result flows directly into the CRM. No copy-paste, no CSV imports, no double entry.
AI Research Engine. Each new lead is enriched automatically. A real-time research engine pulls current information on candidates and companies: what they're hiring for, recent news, public profile signals. An AI scoring engine evaluates every profile against the firm's ICP. Matches against open roles surface automatically in the pipeline view.
CV Scanning. When a candidate uploads a CV, AI extracts skills, experience, and qualifications into a structured profile within seconds. Recruiters never type a resume into a database again.
AI Mock Interview. Before a recruiter spends 30 minutes on a screening call, the candidate can complete an AI mock interview. The system asks role-specific questions, evaluates responses, and produces a structured report. Recruiters only invest live time in candidates the AI has already qualified.
Recruiter Call Prep. This is the bridge between AI work and human work. Before a recruiter dials, the workstation surfaces three things on one screen: the verified strengths the candidate brings to the role, the gaps the AI identified against the job requirements, and a list of suggested questions tailored to that specific candidate and that specific role. Recruiters walk into every conversation already prepared. The AI does the analysis. The recruiter does the talking.
CRM. Candidates, companies, job openings, applications, all in one place, all visible to the whole team across multiple regions. Pipeline analytics show where each candidate is, what is blocked, and where the next action lives.
Email Operations. Full inbox and campaign engine, integrated with the CRM. Every email is tied to a candidate or company record. Templates support variable injection from CRM fields. Open and reply tracking feeds back into pipeline analytics.
Architecture choices we made for them
- Desktop-grade workstation, not a web app. Recruiters work from one machine across a full day, not in a browser tab between Slack notifications.
- Shared cloud database. The team sees the same pipeline in real time across multiple recruiters and multiple regions.
- Encrypted local cache. Sensitive candidate data stays protected on every machine, with API credentials managed at the OS-keyring level.
- Hardened production infrastructure. Deployed on a security-hardened server with full auditing, automated backups, and continuous monitoring.
Cost economics back the architecture. SHRM's benchmarking research pegs the US average cost-per-hire at $4,129 per non-executive role (SHRM Benchmarking Report), and analyses aggregating LinkedIn Talent Solutions data find AI-enabled recruitment workflows cutting that figure by roughly 27% through automated sourcing, screening, scheduling, and pipeline updates (Senseloaf AI on LinkedIn-research cost reductions). People Up's workstation rolls every one of those workflow steps into one place.
Status
Live in production. Currently in active use by the People Up recruitment team across multiple European markets and Canada. https://peopleup.llc
Most "AI in recruitment" pitches are a single feature bolted onto a generic CRM. A CV parser. A mock-interview plugin. A research tool. They look impressive in demos and fall apart in real workflows because the AI does not know the pipeline and the pipeline does not share data with the AI. Most agencies bolt AI onto your workflow. We rebuild the workflow around AI. We built that for People Up. We would build it for you.