Weekly editorial brief ready. 📌 Hot themes (×3): AI Act audit reporting (18 CRM + 7 tickets), integration with enterprise management systems (12 CRM + 4 tickets), open core vs managed model (9 CRM).
AGENT · OVERNIGHT EDITORIAL BRIEF
The marketing editorial plan starts from the real conversations of the day before.
Overnight Editorial Brief works overnight. It reads the CRM (commercial conversations), the ticketing system (customer questions and complaints), mentions on social channels. It extracts the hot themes, the recurring questions, the topics the audience is raising. On Monday morning, the content manager opens the work channel and finds the editorial brief already ready.
Overnight Editorial Brief at work.
❓ Unanswered questions (×2): how to choose the LLM model per sector, difference self-host vs managed (article 14 months old). 🔺 Emerging: EU data sovereignty (12 conversations this week vs 0 in the previous 4).
Perfect. 'How to choose the LLM model' becomes the priority article. 'EU data sovereignty' I'll propose as a CEO LinkedIn post for Friday.
Planning recorded. Have a good editorial week.
Why it exists.
The marketing editorial plan is often built on assumptions: what we think interests the audience, which topics are hot in the sector, which recurring questions would deserve an article. The assumptions are often accurate, but they are still assumptions. The real conversations with customers — in the CRM, in tickets, on social channels — contain the signal of what the audience is actually asking, complaining about, looking for.
How it prepares the overnight brief.
The operational problem is not the absence of data. The systems are there. It is filtering: the systems are separate, the volume is high, no one has time to summarise before Monday's editorial meeting. Overnight Editorial Brief solves this step. The brief arrives before the meeting, built on the week's real data, with theme frequencies and the source for each.
The decision stays with the content manager.
The agent lays out the brief with real data: hot themes, unanswered questions, articles to update, emerging topics. The decision on what becomes an article, LinkedIn post or FAQ stays with the content manager.
Content manager, head of marketing, content writer.
Content manager
Grounds the weekly editorial plan on real conversation data, not on assumptions. Editorial production converges on topics the audience is actually searching for, with frequencies and sources to back planning decisions.
Cross-channel · 7d 1,420 mentions
Site chat NPS 7.2
Reviews top theme: shipping
Social · IG sentiment +0.4
Brief to the brand owner
Head of marketing
Sees, in a structured way, the rhythm of customer conversations and the alignment between editorial production and real demand. The editorial effectiveness measure becomes evidence-based, not perception-based.
Brief 07:00 12 active segments
Gold segment +8% engagement
Silver segment redemption -3%
Anomaly · campaign 14 discarded
Monday action map
Content writer
Receives a brief with already-qualified topics, frequencies, concrete conversation examples. Reclaims the time spent on topic research and concentrates capacity on writing.
RB RevOps Brief
Idritech SpA silent 8d · proposal expiring
Nexar Group proposal sent · no reply
Aludra Srl signature expected this week
Monday morning at six. The brief is ready before the editorial meeting.
The agent is scheduled at four in the morning every Monday.
For the content manager of a regulated B2B SaaS company, the Monday afternoon editorial meeting plans the week. Overnight Editorial Brief is scheduled at four in the morning every Monday. It reads the CRM for the previous week's commercial conversations, the ticketing system for opened and closed tickets, and social channel mentions.
Hot themes, unanswered questions, emerging topics.
It extracts the week's hot themes (AI Act audit reporting, integration with enterprise management systems, open core model), unanswered questions (questions the audience asks that no existing content covers), and emerging topics (words or concepts that appeared this week but not in previous weeks). The brief reaches the marketing Slack channel at six in the morning in four structured blocks.
The content manager identifies the week's priority.
The content manager reviews the brief, identifies the week's priority, and proposes the articles. The editorial plan is grounded in data, not in intuitions. The event — themes extracted, content manager decisions, articles planned — stays in the audit registry for editorial effectiveness analysis.
Configuration and technical resources.
The Overnight Editorial Brief rules are declarative. The customer's marketing team defines the source systems to read, the editorial corpus for comparison (blog, site, historical posts), the frequency thresholds to declare a theme "hot", the brief format. The rules live in the customer's repository, versioned.
CRM integration is native for HubSpot. For other CRMs (Salesforce, Microsoft Dynamics), integration is delivered during the project. For ticketing systems (Zendesk, Freshdesk, proprietary systems) and social channels (LinkedIn API, social listening tools), integration is delivered during the project via dedicated adapters.
- Language
- TypeScript (Node.js)
- LLM model
- customer's choice: Anthropic, OpenAI, Mistral, open source models hosted internally, AWS Bedrock for a private model
- Built-in controls used
- pii-detector, tool-rate-limit, message-length-limit
- Native delivery channels
- Slack, Telegram, WhatsApp, OpenAI-compatible HTTP
- Scheduling
- configurable per instance (typical Monday night/morning)
- CRM integration
- HubSpot native; other CRMs (Salesforce, Dynamics) in delivery
- Ticketing integration
- dedicated adapter delivered during the project (Zendesk, Freshdesk, proprietary systems)
- Social channel and social listening integration
- dedicated adapter delivered during the project (LinkedIn API, X API, Sprinklr, Khoros, Brand24)
- Memory
- persistent per instance, pgvector + PostgreSQL FTS on the customer's editorial corpus
- Registry
- immutable, queryable with a standard SQL client
Frequently asked questions about the agent.
No. The agent lays out the brief with real data: hot themes, unanswered questions, articles to update, emerging topics. The decision on what becomes an article, LinkedIn post or FAQ stays with the content manager. The agent prepares the signal; the human turns it into the plan.
The CRM is natively covered for HubSpot. For other CRMs (Salesforce, Dynamics), ticketing systems (Zendesk, Freshdesk) and social channels (LinkedIn, X, social listening tools), integration is delivered during the project by the Exelab team. The commercial discovery qualifies the integration patterns for the real case.
On the improvement in editorial plan quality and the performance of published content. Articles that answer unanswered questions identified by the agent typically show engagement rates above the customer's editorial corpus average. The baseline is established during discovery.
When conversation volume is not enough to extract statistically significant themes, the brief says so openly. The content manager receives a note — "low volume this week, insufficient themes" — rather than a brief with fictional themes. It is a low-activity signal, not an agent error.
From a 30-minute conversation to the squad in production.
A 30-45 minute conversation to understand how Overnight Editorial Brief would configure to the customer's case. Which source systems, which editorial corpus for comparison, which brief cadence.