AI negotiations in Work App: designing automated offers

Negotiating rates at scale is the biggest bottleneck in high-volume hiring. We designed AI negotiations in Work App to handle candidate price discussions in the background, keeping clients in complete control of their budget.

In our previous articles, we explored bulk hiring and auto-hiring in Work App. This third chapter covers how we built AI-driven price negotiations and autonomous workflow controls into the platform.

Rigid screening filters work well when hiring requirements are straightforward, but real-world hiring often comes down to rate negotiations and availability. When posting high-volume jobs, manually negotiating rates with dozens of individual applicants creates a massive bottleneck. AI negotiations allow clients to hand off dynamic bid negotiations to an automated assistant while keeping budget caps strictly intact. The core design challenge was giving the AI enough flexibility to secure competitive bids without taking control away from the client or frustrating applicants with robotic, opaque interactions.

Imagine you are hiring thirty UGC (user generated content) creators to produce short-form video reviews for a global product launch. Applicants submit a wide range of bids based on their reach, portfolio quality, and turn-around speed. Instead of spending hours messaging candidates individually to discuss rates or make counter-offers, you need a system that handles those negotiations on your behalf in the background, staying strictly within your target budget.

Clients can enable AI negotiations during initial job creation or turn them on at any point during the hiring process for an existing post. Once active, applicants receive interactive offer cards in their chat thread directly from the AI assistant in real time. The AI manages counter-offers based on simple upfront controls set by the client, including target savings goals and conversation tone.

For more advanced workflows, external AI agents can also be authorized via external commands to post jobs, negotiate rates, manage project progress, and trigger escrow payouts using a pre-funded budget. The client maintains complete oversight and can step in, modify settings, or pause the system at any moment.

We couldn’t let an AI assistant negotiate without clear boundary limits set by the client. However, forcing clients to configure complex rules or write detailed instructions would create heavy setup friction and defeat the point of background automation.

High-level negotiation presets

To simplify setup, we organized the controls into clear, high-level parameters. Clients select a negotiation strength target, choosing between low, medium, or high savings goals, and set a conversational tone ranging from friendly to direct or persuasive. Clients also define strict message volume limits to control how many counter-offers the AI can send across the campaign.

Clients get notified when it’s time to top up negotiation message balances if a high-volume campaign runs low. Negotiation strength and tone parameters can also be modified on active jobs without interrupting ongoing applicant conversations. A master toggle on the job settings screen allows clients to pause or turn off AI negotiations instantly.

Dynamic negotiations require clear options for applicants and real-time visibility for clients. If the negotiation flow feels rigid, applicants walk away. And if clients cannot see active offers at a glance, they lose trust in the background process.

Interactive offer cards

Rather than auto-accepting bids or forcing applicants into complex forms, we designed interactive offer cards directly inside the applicant’s chat thread. Applicants can accept the proposed rate in a single tap, dismiss the offer, or edit their bid to submit a new counter-offer.

Status banner and offers page

On the client side, we kept the main job view clean by displaying a prominent status banner at the top of the job page. Drilling down opens an expanded offers page, giving clients full visibility into active conversations, candidate counter-bids, and total savings achieved so far, along with individual controls to override or stop specific negotiations.

Beyond in-app price negotiations, we wanted to allow external AI agents to handle entire hiring workflows, from posting jobs and negotiating rates to managing deliverables and approving payouts. The design challenge was granting external agents the operational freedom to run complex hiring workflows independently while building strict financial safety nets so clients never risk runaway spending.

Pre-funded escrow budgets

We solved this by giving agents their own capped escrow budget instead of open access to the client’s account balance. External agents can only post jobs, negotiate rates, and trigger payouts up to the exact balance allocated to their dedicated escrow budget. If an agent reaches its budget limit, all automated actions pause immediately.

Clients can also disconnect external AI agents with a single tap from the settings screen without affecting active escrow jobs or current freelancer contracts. This gives external agents full operational freedom to run hiring workflows while guaranteeing absolute financial safety for the client’s account.

Over this three-part series, we have explored how Work App evolved from manual batch actions to background auto-hiring, and finally to intelligent AI negotiations and agent workflows. By pairing background automation with granular controls, pre-funded escrow protections, and transparent status tracking, Work App enables clients to scale their hiring operations while maintaining complete control over their budget.