Job.com: What We Are Delivering

Oct 5, 2026

Product vision

On job.com, jobs apply to you. Every candidate gets an AI agent that knows their license, skills, pay goal, and start date, watches every open order, and puts them forward in minutes with a complete packet. Behind the agent is a staffing company run by software: the platform does the matching, pricing, packaging, and follow-up, and people do the parts that need people. We start with healthcare staffing because it is the hardest case; a platform that works there works anywhere.

What the board sees on

A live demo on real orders, real candidates, and real pricing. Nothing on stage depends on a customer or a connection we do not have. Alton opens with the story and the problem in numbers (1,028 live orders, the lowest vendor tier with every MSP, 2 hours per packet, 4 days per resume, pricing errors) and closes with the roadmap. In between, I demo the product:

1. Orders in. Live orders flowing from PCRecruiter into Job Desk, the recruiters' new workspace.
2. Matching. Ranked candidates for open roles.
3. Candidate onboarding. A candidate signs up and consults with the AI agent who builds their profile and shows matched jobs on the first screen.
4. Candidate-first view. One candidate and the ten orders they fit, ranked by bill rate and their preferences.
5. The AI agent reaches out. It contacts the candidate about a strong match; a yes is their consent to be represented, and they land in the recruiter's queue.
6. Pricing and packet. The pay package priced at 16.5% margin and the MSP submission packet, generated in ten seconds.
7. Scoreboard. The KPIs the system is tracked and tuned against: time from order to first submission by MSP, complete candidate profiles, and the fill rate that raises Job.com's vendor tier.

The first of these ships before Austin: the pay package calculator goes to Julia and Anna this week, removing the two-hour approval wait and the pricing errors.

Where this goes

Everything on stage in Austin is the first version of the platform that runs Job.com for the next two years. By the end of December the MSP business runs on it: the AI agent finds and screens candidates, orders arrive and are matched automatically, pricing and packets take seconds, and the agent stays with each candidate through the contract and into the next contract. From there the same engine takes on the rest of a staffing business one job at a time, adds verticals as configuration rather than rebuilds, and opens the path to direct employer contracts, until Job.com is a staffing company run by software and measured on numbers no staffing peer can match.

Proof by Dec 31: order to first submission in hours, not days; zero pricing errors; complete profiles growing every week; placements per recruiter up against the September baseline; vendor tier improved.

Who does what

• Alton owns the business: the operating plan, go-to-market, verticals, hiring and budget, and the board story. He also builds the AI agent's harness.
• Stan owns product and engineering: what gets built, in what order, and when it ships, including the matching engine, Job Desk, the candidate portal, and the AI agent. He writes the assignments and runs the weekly product review.
• Miriam designs frontend: marketing site, candidate screens, and job desk - one deliverable at a time, with a date.
• Max builds the PCRecruiter bridge, then the VMS connections, one deliverable at a time, with a date.

The full plan, with the month-by-month sequence, the roles, and the open decisions, is here: Job.com Platform Plan