Fountain's Orchestra Builds AI Agents for Frontline Hiring and Shifts
Fountain released Orchestra on November 17, 2026, letting employers build AI agents for hiring and scheduling in plain language, with supervisory checks and humans keeping hiring decisions.

Why it matters
- Fountain released Orchestra on November 17, 2026, for building AI agents that handle hiring and scheduling tasks.
- More than 500 frontline employers across logistics, retail, restaurants, healthcare and staffing use Fountain's software.
- A supervisory check compares each agent's actions and results against its goal before work is marked complete.
- Fountain runs more than 180 agents internally across engineering, customer success, sales and finance.
- General availability is planned for early 2027 with usage-based pricing tied to proven completed agent work.
Fountain released Orchestra on November 17, 2026, a platform that lets employers build AI agents for hiring and scheduling by describing the work in plain language. The agents draw on data inside Fountain's software, call workers and applicants, and report whether each assigned task was completed.
The stakes are straightforward. More than 500 frontline employers across logistics, retail, restaurants, healthcare and staffing already run Fountain's hiring and workforce software. For them, a 2 a.m. dropped shift or an applicant who goes quiet over the weekend costs money and time. Orchestra aims to close those gaps automatically, while leaving the decisions that matter to people.
Fountain's chief product and technology officer, Salim Jernite, framed the design problem in blunt terms: "An agent you can't check is a risk. An agent that tells you when it's wrong is one you can put to work."
What can Orchestra's agents actually do?
Orchestra targets recurring frontline work. Agents are shared across a team rather than confined to one user. Initial use cases include:
- Covering open shifts
- Re-engaging past applicants before an employer pays for job advertisements
- Preparing new hires for their first day
- Following up with applicants
The setup process starts with a description. An operator describes a job to hand off, or specifies a goal such as reducing time to hire, and asks Orchestra to propose an agent. Orchestra presents the specifications and requires the user to validate the design before building anything.
Once approved, an agent can run three ways: on demand, on a schedule, or in response to an event such as a dropped shift or an opened position.
How does Fountain verify an agent finished the job?
Every run follows the same structure. The agent sets a goal and a plan. Then a separate supervisory check compares both the result and the agent's actual actions against that goal before the task can be marked complete.
This mechanism separates activity from completion. An agent must be checked against its assignment, not simply report that it ran. Work that falls short gets flagged with reasons for human review, giving the team a record of what was finished and what still needs a decision.
The practical effect: a shift dropped at 2 a.m. or an applicant who goes quiet over the weekend can be addressed when it happens, with unresolved items left for a person to settle.
What contact safeguards are built in?
Orchestra-built agents operate only within the permissions of the person running them, and every action is recorded in plain language. Fountain says agents that call workers or applicants check consent, calling hours and daily contact limits on every call. The agents also disclose that the caller is an AI and honor opt-outs.
For hiring, those contact rules sit alongside hard limits on decision-making. According to Fountain, agents contacting people use objective criteria established in advance, refuse discriminatory requests, and leave hiring decisions to humans. The design separates automated outreach and task execution from the employment decisions that remain with managers and hiring teams.
Fountain is also running the approach on itself. The company reports operating more than 180 agents across engineering, customer success, sales and marketing, and finance, with employees reviewing the agents' work and making decisions. That internal deployment provides the operational background for Orchestra's emphasis on supervisory checks and reviewable records.
How does Orchestra fit with Cue?
Orchestra extends Cue, which Fountain introduced in April 2026 to execute sourcing, screening, candidate scheduling and shift-gap work. Fountain's platform documentation describes Cue as the interface coordinating automated workflows across hiring, onboarding and shift scheduling, with built-in agents handling tasks such as applicant screening and onboarding. Fountain reports that customers have asked Cue more than 80,000 questions since launch.
Fountain's materials draw a clear line between the two levels of control. Cue coordinates the platform's established workflows. Orchestra lets a team define an assignment or goal, validate the proposed agent, and share it with colleagues. That moves the new product beyond issuing requests to an existing assistant and into configuring agents around a team's own recurring responsibilities.
What does it cost and when is it available?
Orchestra is available at no charge to select customers within existing Fountain accounts right now. Fountain is targeting general availability for early 2027, when it plans to charge for the work each agent performs and let customers control how much work runs.
Under that plan, employers would pay for completed agent work, while people remain responsible for hiring decisions and for reviewing tasks that do not meet their goals. Usage-based pricing in early 2027 depends on proving agents complete useful work. The wider rollout, in other words, turns on whether Orchestra's supervisory checks can demonstrate enough completed, useful work to support that pricing model across larger frontline teams.
For frontline employers, the bet is that verified agent labor becomes a line item they can scale. For Fountain, it's whether trust in the check — not the agent — becomes the product.
Source: unite.ai
More from Elena Vasquez
Show full bio
Market editor covering technology and product news in artificial intelligence and machine learning at AI In Context.
708 articles
Related articles
- OpenAI Open-Sources Symphony, a Spec That Turns Linear Into an Agent Orchestrator
- OpenAI Says Agents Have Replaced Chatbots as Its Default Work Tool
- Restate raises $20M Series A to challenge Temporal in agent infrastructure
- Ceipal Launches Ora AI Agents, Says 650 Customers Already Deployed
- Australian firm opens the books on five months of AI agent controls