Monday morning opportunity brief. 11 qualified across 540 customers. Arcadia Ltd: intensive Analytics Plus for 8 weeks on Base plan, candidate Premium, email draft for Mark R. Meridian Logistics: integration with system Y activated, candidate for related product Z, draft for Sarah T. NovaBuild plc: 4 new users in 6 weeks, plan limit reached, candidate Enterprise, draft for Luke F. → 8 more in thread.
AGENT · CROSSSELL OPPORTUNITY DETECTOR
Upsell and cross-sell opportunities arrive already qualified for the sales team.
CrossSell Opportunity Detector reads weekly the adoption signals on the customer's product, recognises patterns compatible with an upsell or cross-sell opportunity, and prepares for the assigned sales rep a brief with a pre-filled email draft. The sales rep decides, the agent lays out.
CrossSell Opportunity Detector at work.
Perfect. Mark and Sarah proceed with the drafts. On NovaBuild we wait for the call scheduled Friday.
Decisions recorded in the CRM. Opportunities updated with status.
Why it exists.
Upsell and cross-sell are the two multipliers of the customer relationship: additional products to customers who already know and use the company. Mid-to-large companies typically have enough data to identify ready customers — customer success systems, product adoption data, usage behaviours — but the signal flow to the sales rep is not structured. The result is that most opportunities are lost or materialise late.
How it works each week.
CrossSell Opportunity Detector activates weekly. It reads the adoption signals on active customers, compares against the opportunity patterns configured by the customer success or sales operations team, and identifies customers showing a pattern compatible with upsell or cross-sell. For each opportunity it prepares for the assigned sales rep a brief with the customer context and a pre-filled email draft.
The decision stays with the sales rep.
The decision stays with the sales rep. The agent does not write to the customer autonomously, does not modify the CRM without approval, does not close opportunities without the sales rep.
The corporate teams that turn adoption data into commercial action.
Commercial director
Sees the weekly upsell and cross-sell opportunity flow in a structured way, separated by assigned sales rep. Opportunities become a real picture rather than an anecdotal perception. Conversion rates become measurable over time.
RB RevOps Brief
Idritech SpA silent 8d · proposal expiring
Nexar Group proposal sent · no reply
Aludra Srl signature expected this week
Customer success lead
Sees the commercial value generated by their own adoption signals. The data chain adoption → signal → sales rep → customer becomes measurable and improvable: opportunity rules are refined based on real outcomes.
87 Tone too formal — try a direct opener
71 Long message: summary in the first 2 lines
94 Excellent — consistent with the customer history
Assigned sales rep
Reclaims the time spent preparing the initial email. The brief arrives ready on the work channel at the start of the week, with context and draft. The sales rep decides whether to proceed and personalises the message.
CustomerSignalRisk
Idritech SpA Silent 12 days · 3 emails unanswered high
Nexar Group Proposal expired · no extension med
Orbis Srl Meeting cancelled 2× med
Proxa Corp Active contact · reply < 24h low
Eleven qualified opportunities, delivered every Monday morning.
Monday at seven, 540 customers analysed.
For the commercial director of an EU-regulated B2B SaaS company, Monday morning starts with the weekly opportunity brief. CrossSell Opportunity Detector is scheduled at seven every Monday. The agent reads the customer success system for the 540 active customers from the previous week. It compares usage patterns against the eight configured opportunity rules. It identifies 23 candidate opportunities. It filters by minimum confidence threshold: at least three compatible signals in the last four weeks. Eleven qualified opportunities remain.
For each, context and pre-filled email draft.
For each opportunity, the agent composes the brief for the assigned sales rep. For the customer on the base plan with intensive use of the premium feature over recent weeks, a brief with an email draft proposing a premium plan evaluation call. For the customer who just activated an integration with a third-party system, a brief with an email draft proposing the related product. For the customer in internal growth phase with the current plan user limit nearly reached, a brief with an email draft proposing an enterprise plan planning call.
Seven confirm. Three rework. One postpones.
The eleven opportunities arrive on the Slack channels of the respective sales reps. Seven confirm the draft with minor changes. Three substantially rework it before sending. One postpones for customer-specific context. All decisions stay in the runtime audit registry, readable by the commercial director with a standard SQL client for effectiveness analysis.
Declarative rules from the customer's customer success or sales operations team.
The opportunity rules of CrossSell Opportunity Detector are declarative. The customer's customer success or sales operations team defines in a readable format the opportunity patterns — what in adoption data indicates an upsell candidate, what indicates a cross-sell candidate, which signals are needed in combination — the confidence thresholds for qualification, and the format of the email draft to generate. The rules live in the customer's repository, versioned, validated at agent startup.
- 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, credential-detector, topic-guardrail, tool-rate-limit
- Native delivery channels
- Slack, Telegram, WhatsApp, OpenAI-compatible HTTP
- Scheduling
- configurable per instance (typical weekly, Monday morning)
- CRM integrations
- HubSpot native (contacts, deals, tickets, meetings, tasks, notes, company, email)
- Other CRMs
- Salesforce, Microsoft Dynamics, sector-specific systems: integration delivered during the project by the Exelab team
- Customer success / product analytics system
- dedicated adapter built during delivery (system varies by customer)
- Memory
- persistent per instance, pgvector + PostgreSQL FTS on usage patterns
- Registry
- immutable, queryable with a standard SQL client
How CrossSell Opportunity Detector works in detail.
No. The sales rep stays the owner of the customer communication and the send decision. The agent prepares the brief and a pre-filled email draft; the sales rep reviews the context, modifies the draft, and decides whether to send today or postpone. The cardinal principle: the agent lays out the sales rep's first step.
The customer success system or product analytics system is the cardinal system — it tracks usage patterns. The CRM is the second, for commercial interactions. HubSpot is available natively. For other systems — Salesforce, Microsoft Dynamics, sector-specific management systems, proprietary customer success systems — integration is delivered during the project by the Exelab team.
Every identified opportunity is inspectable, and the outcome of every rule is traced separately. When an opportunity is flagged as not relevant for the customer's specific case, the customer success or sales operations team modifies the declarative rule, tests it in the development environment, and promotes it to production. Rule improvement stays inside the customer's team.
The value of CrossSell Opportunity Detector is measured on the percentage of identified opportunities that materialise as upsell or cross-sell within a configurable number of weeks. The pre-agent baseline is established during discovery. The first weeks of operation compare the flagged opportunity flow against the baseline. CrossSell Opportunity Detector is one of the agents the customer can choose to build for their commercial team.
From a 30-minute conversation to the squad in production.
A 30-45 minute conversation to understand how CrossSell Opportunity Detector would configure to the customer's case. Customer success systems, opportunity patterns, delivery frequency.