What if a job req could start as a sentence, and end as a pipeline? Now it can, in Amazon Quick, with SeekOut. Quick is the AI companion built for work, an enterprise-grade AI assistant that handles real work across your systems. Your work happens in a dozen tools, and Quick connects them all. One of those is SeekOut.

Connect SeekOut once, over the Model Context Protocol (MCP), and your whole team can recruit in plain language without leaving Quick. Describe the role you are hiring for, and SeekOut does the recruiting work behind every answer. The 3 minute demo below follows one Staff Machine Learning Engineer search from the first sentence to a pipeline.

Recruiting, inside Amazon Quick

An assistant that only knows about recruiting can describe what a good search might look like. With SeekOut connected, Amazon Quick runs the search. It works against SeekOut’s data, ranks real candidates against your own criteria, creates SeekOut workspaces, and retrieves verified contact details, all from the same chat. Answers drawn from SeekOut show it as the source, and the same shortlist is one optional click away in the native SeekOut app whenever you want depth.

The whole flow, in a single conversation

Here is the exact flow in the demo, for a Staff Machine Learning Engineer on an AI agents team, with production LLM experience, strong Python, and distributed systems depth. Every count below is what this run returned or what we asked for, not a limit.

• Start with the role, not a search string. Before it looks at a single profile, SeekOut separates what the role truly requires from what is only nice to have, names the titles people actually carry, and lines up the target companies.
• See where the role is winnable. It compares the markets you ask about (in this example Seattle, the Bay Area, New York, and Toronto) on pool size, compensation, competition, and risk, so you walk into the kickoff with a plan.
• Run the search. In this example, 566 people match across Seattle and the Bay Area, and each one arrives with the signal behind it, not just a name.
• Get a ranked shortlist. Ask it to evaluate the top candidates, and you get a hiring-manager-ready ranking with the evidence for each person, the concerns or gaps, and what to ask on the first call.
• Approve the action before it runs. Ask it to create a SeekOut workspace for the shortlist, and Amazon Quick names the action it wants to run, then waits for your yes.
• Open the same shortlist in SeekOut, if you want to. It hands you a link to the native SeekOut app: the 25 we asked for, scored against your own criteria, with the whole pool from this run behind them, 575 profiles in this example.
• Get contacts, with the SeekOut credits confirmed first. Ask for contact details, and it tells you exactly how many SeekOut credits it will spend, then waits. In this example, we take just the top seven it evaluated, and the verified emails come back.
• Draft outreach that reads like a recruiter wrote it. Here, we ask for ten drafts, each one built from what that person has actually shipped.
• Widen the funnel, not the bar. Ask how to expand the pool without lowering the bar, and you get multiple strategies with the tradeoffs attached, from alternate titles and adjacent skills to feeder companies and nearby markets.
• Take a second look at your own ATS. Run the same role against your ATS, and the people who already applied to you come back ranked, with the reason each one is worth revisiting.

That is the whole flow, from a job req to a pipeline, without leaving the chat.

Why it matters

• It works in the AI your company already approved. There is no new tool to evaluate or roll out. If your team already asks Amazon Quick for research and business insights, recruiting now happens in the same place.
• Set up once, used by the whole team. The connector is created in the Amazon Quick console, and each recruiter signs in with their own SeekOut account.
• Plain language in, a pipeline out. Nobody learns a query syntax or builds a Boolean string. You describe the role.
• Evidence, not a black box. Each match carries the signal behind it, and the shortlist comes with the evidence, the gaps, and the questions to ask, so you can defend it to a hiring manager.
• Your approval, where it matters. Before it creates a workspace or retrieves contacts, Amazon Quick names the action and waits, and it tells you how many SeekOut credits a step will use before it spends any.
• Your ATS becomes a source again. The people who already applied to you get a second look, ranked against the role you are hiring for today.

These eight are just the start

Every prompt in the demo is in the SeekOut prompt pack, word for word, ready to paste into Amazon Quick. It is one growing library, built for recruiters working inside Amazon Quick: the eight-step workflow from this demo, plus prompts for recruiter productivity, TA leadership intelligence, and execution, from building a competitive talent map to rediscovering silver medalists. No sign up, no gate.

Get the prompts

Get started

It only has to be set up once. You need an Amazon Quick Enterprise subscription with connector permissions, and a SeekOut account to sign in with.

1. In the Amazon Quick console, open Connectors.
2. Select Create for your team, and choose Model Context Protocol (MCP).
3. Name it, then paste the SeekOut endpoint: https://seekout-search-mcp.seekout.io/tools
4. Select User authentication (OAuth), and sign in to SeekOut.

Then just ask. Start with a role you are hiring for, or open the prompt pack and run the eight from the demo. New to SeekOut? Start a free trial and bring your own roles.

Start a free SeekOut trial

Build the way you hire. With Amazon Quick and SeekOut.