Deploying AI Agents

Deploying Questera’s AI agents and configuring their objectives. An agent in Questera is an autonomous AI module focused on a specific aspect of customer engagement or marketing. Each agent is configured with a clear goal, a set of rules/constraints, and a toolbox of abilities it can use. You can create custom agents or leverage Questera’s pre-configured agents (such as an email marketing agent, a retargeting ads agent, a segmentation agent, etc.) for common growth tasks.

Agent Goals

The goal defines what success looks like for that agent. It could be maximizing a KPI or triggering a specific outcome. For example, one agent’s goal might be “increase 7-day user retention by 20%” while another’s goal is “maximize cart-to-purchase conversion rate.” In practice, goals are tied to conversion events or metrics from the ingested data. During setup, teams define an objective with parameters such as the target event (e.g. purchase_completed) and possibly a numeric target or forecast. Agents can even optimize for multiple success events concurrently by weighting them, enabling sophisticated multi-objective optimization. These goals inform the agent’s decision-making and what it treats as a “reward” when evaluating outcomes.

Abilities & Modular Services

Abilities are the actions and services an agent can execute. Questera provides a library of modular, pluggable services (integrations) that agents can leverage – akin to skills in the agent’s toolkit. When configuring an agent, you choose which abilities to grant it. Examples of abilities include: sending an email or SMS via your ESP (Email Service Provider), showing an in-app notification, adding a user to a retargeting ad campaign on Google/Facebook, generating a personalized product recommendation, or invoking a discount offer. Each ability is powered by an underlying integration or API. For instance, the “email send” ability might connect to Braze or SendGrid; the “ads retargeting” ability connects to Google Ads or Meta; a “segment user” ability could trigger a CRM update, etc.

Questera’s platform has 75+ such integrations available, meaning agents can act across channels and systems seamlessly. Technically, these abilities are implemented as modular services – the agent uses function calls to external APIs or internal microservices to perform tasks (What is AI Agent Orchestration? | IBM). This modular design makes the system extensible: new abilities can be added without redesigning the core platform, and each agent can be tailored with the exact capabilities needed for its mission.

Once agents are deployed, Questera’s orchestration engine takes over to coordinate their activities and deliver personalized campaigns. Think of the orchestration engine as the central “AI conductor” that manages when and how each agent engages with a customer). Rather than running fixed drip campaigns, Questera continuously orchestrates a dynamic, adaptive interaction flow for each user based on live data.

Rules & Constraints

While agents are autonomous, marketers can set business rules or constraints to guide their behavior. Rules ensure the AI stays within brand guidelines and ethical boundaries. Examples include frequency caps (e.g. “don’t send more than 2 messages per day to a user”), scheduling constraints (“avoid contacting users before 8 AM local time”), channel preferences, budget limits for spend, or compliance rules (GDPR/CCPA considerations). Rules essentially act as guardrails – they are the “do’s and don’ts” the agent must obey while pursuing its goals. In Questera’s configuration interface, setting core instructions means specifying such rules and parameters for the agent (Questera - Agentic Customer Engagement Platform). This ensures the AI’s autonomous actions align with business strategy and customer experience standards.

Agent Configuration Schema

When deploying an agent, you define its configuration in a structured way. Here is a conceptual schema for an agent configuration:

{

"agent_name": "RetentionBooster",

"goal": {

"objective": "increase_retention",

"success_event": "app_open_7_days",

"target_metric": "DAU_week+1",

"target_increase": 0.20

},

"rules": {

"max_messages_per_day": 2,

"allowed_channels": ["push_notification", "email"],

"quiet_hours": {"start": "21:00", "end": "08:00"}

},

"abilities": [

"send_email",

"push_notification",

"offer_dynamic_discount"

]

}

In this example, the agent “RetentionBooster” is configured to increase user retention (defined as a higher 7-day active rate). It has rules limiting contact frequency and timing, and it has three abilities (email, push, and discount offering) enabled. In Questera’s platform UI, this would correspond to entering the agent’s goal parameters, setting toggles or values for various rules, and selecting integrations (abilities) from a list. Once configured, the agent is deployed into Questera’s AI workforce.