> For the complete documentation index, see [llms.txt](https://docs.questera.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.questera.ai/questera-ai-platform/ai-agents/agent-orchestration.md).
# Agent Orchestration
Agent orchestration is how different AI agents coordinate in real time — across channels, stages, and user segments — to deliver personalized, adaptive, and compounding growth.
Instead of static flows or campaign calendars, Questera’s agents:
* Make decisions in context
* Collaborate with each other
* Adapt to user behavior
* Optimize based on feedback
**How It Works Together**
* The Retrieval & Memory Layer pulls long-term preferences and recent interactions, like RAG for personalization.
* The Agentic Layer decomposes the decision:\
→ "What’s the user’s real intent?"\
→ "Where are they in the funnel?"\
→ "What message will resonate?"\
→ "What’s the best timing + channel?"\
→ "What visuals and tone will work best?"
* Each agent specializes (like a team of marketing strategists).
* A Coordinator/Router Agent can oversee arbitration when multiple actions are viable (e.g., upsell vs nurture).
* The Feedback Agent closes the loop using performance signals (CTR, revenue, churn) to reinforce and fine-tune matchmaking over time.
## Agent Roles in Questera
### Specialists (Own the “What”)
These agents are domain experts, each responsible for a key stage or growth outcome:
| Specialist Agent | Focus | What it Does |
| ------------------ | ---------------------------- | ------------------------------------------------------------- |
| Activation Agent | Onboarding & habit formation | Detects drop-offs, prompts key actions, builds early momentum |
| Churn Agent | Retention | Spots usage drops, delivers re-engagement nudges |
| Expansion Agent | Monetization & upsells | Identifies upsell triggers, recommends upgrades or add-ons |
| Reactivation Agent | Winbacks | Targets dormant users with personalized recovery campaigns |
### Routers (Own the “How” and “When”)
These agents optimize delivery:
| Routing Agent | Role | What it Does |
| -------------- | ---------------------- | --------------------------------------------------------------------------- |
| Channel Router | Channel + format | Chooses whether to deliver via email, push, in-app, or SMS |
| Timing Router | Temporal intelligence | Determines best send time based on user behavior & context |
| Tone Styler | Communication strategy | Picks tone & style based on user segment, past behavior, and outcome intent |
### Coordinators (Own the “Why Now”)
These meta-level agents resolve conflicts, manage orchestration, and control pacing:
| Coordination Agent | Role | What it Does |
| ------------------- | ------------------------ | --------------------------------------------------------------------------- |
| Journey Coordinator | Priority resolution | Decides which agent takes action when multiple compete |
| Fatigue Manager | Frequency control | Avoids spamming or overwhelming users with too many touchpoints |
| Goal Router | User lifecycle awareness | Aligns actions to macro goals (e.g., don’t upsell if onboarding’s not done) |
Coordinators keep the system coherent — no step on toes, no overfire, always aligned with the user’s stage
### Memory Agents
Purpose: Retain long-term user context across sessions, campaigns, and channels
| Agent Name | Role | Example Task |
| --------------------- | --------------------------------------------------------------- | ----------------------------------------------------- |
| Profile Memory Agent | Stores user traits, historical actions, and channel preferences | “This user responds better to push than email.” |
| Journey History Agent | Tracks past campaigns, outcomes, and learnings | “Avoid reusing the same winback message from Jan ‘25” |
### Insight Agents
Purpose: Analyze behavior, extract signals, and generate optimization hypotheses
| Agent Name | Role | Example Task |
| -------------------------------------------- | --------------------------------------------- | -------------------------------------------------------- |
|
Segmentation Analyst Agent
| Clusters users based on behavior patterns | “Create a segment for high-spending but low-usage users” |
| Attribution Agent | Maps campaign impact to revenue or engagement | “This campaign lifted conversions by 23%” |
### Creative Agents
Purpose: Generate and refine content tailored to context, goal, and tone
| Agent Name | Role | Example Task |
| ----------------------- | ----------------------------------------------- | --------------------------------------------- |
| Message Generator Agent | Writes variant messages for a goal | “Create 3 churn prevention subject lines” |
| Style Aligner Agent | Ensures tone, formatting, and brand consistency | “Rewrite message to match playful brand tone” |
### Experimentation Agents
Purpose: Automate testing, evaluation, and iteration cycles
| Agent Name | Role | Example Task |
| ------------------ | --------------------------------------- | ----------------------------------------- |
| Split-Test Manager | Sets up controlled experiments | “Test 2 onboarding flows across segments” |
| Optimizer Agent | Reallocates traffic to winning variants | “Move 80% to variant B after 3 days” |
### Guardrail Agents
Purpose: Ensure ethical, legal, and brand-safe operation of agents
| Agent Name | Role | Example Task |
| ---------------- | -------------------------------------------- | --------------------------------------------- |
| Compliance Agent | Scans campaigns for GDPR/CAN-SPAM violations | “Flag this message for opt-out link missing.” |
| Ethics Agent | Avoids manipulative or over-aggressive flows | “Throttle upsell for recently churned user” |