Good morning. Customers in switching window for today: 12 identified. Top priority — Customer A (Business 50GB plan, promotion expiring in 18 days + 35% consumption drop): dedicated proposal for Business Plus plan with 15% discount on early renewal. Callback today recommended. Customer B (Consumer base plan + incoming portability signal from competing operator): proposal for plan upgrade with data bonus. Escalation to senior retention within 48 hours.
AGENT · CUSTOMER WIN-BACK
Telco customers in a switching window arrive qualified to the retention team.
Customer Win-Back identifies telco customers in a competitive switching window (promotion expiry, incoming portability signal) and prepares a personalised proposal for the retention team — dedicated offer, structured callback, proactive portability activation.
Customer Win-Back at work.
Confirmed. I'm calling Customer A now. Passing Customer B to the senior.
Decisions recorded in the CRM. Callback scheduled. Regulatory audit registry updated.
Why it exists.
The telco market has a high churn rate. Retention is central to the economic sustainability of operators. Manual win-back management for customers in a switching window is often reactive: the intervention arrives after the portability request, when the customer's decision is already made.
How it works every day.
Customer Win-Back runs on a daily schedule. It identifies customers in a competitive switching window: promotion expiry within the next 60 days, external incoming portability signals, a consumption drop typical of a customer moving away. Based on the customer's profile, it prepares a personalised proposal for the retention team: a tailored offer consistent with the customer's profile, a first check-in call, and where relevant, proactive portability activation on another operator in the group.
The decision stays with the team.
The retention team decides. The agent prepares the context and the proposal; the customer relationship stays with the human team.
The telco functions that change the way they manage retention.
Retention manager
Receives every morning a list of customers in switching windows with the personalised proposal already prepared. The team's time focuses on the retention conversation, not on finding at-risk customers.
Switch window · 7d 240 customers
High-value cluster A callback prio 1
Dedicated offer €8/month · 12 months
Engagement < threshold escalation to operator
Expected win-back rate: 22%
Commercial manager
Monitors the volume of customers in switching windows and the percentage of those who, after the team's intervention, renew their contract. The retention metric becomes manageable ahead of time.
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
Compliance officer
Has the audit trace of retention interventions, inspectable against regulatory obligations on the transparency of commercial offers.
fnol.receive 09:14:22 ALLOW
triage.classify 09:14:25 ALLOW
idd.check 09:14:31 WARN
liquidation.propose 09:15:02 ALLOW
SELECT * FROM audit_log WHERE claim_id = '2024-0847'
A customer in a switching window intercepted before they decide.
120,000 customers analysed, list ready in the morning.
For a telco operator with a portfolio of 120,000 consumer and business customers, the daily Customer Win-Back cycle produces the switching-window list for the retention team. The agent analysed the portfolio overnight: it cross-referenced promotion expiry dates, consumption data over the last 30 days (significant drop relative to the customer's historical average), and incoming portability signals from competing operators.
Profile, history, price sensitivity: a proposal.
For each identified customer, the system prepared a personalised proposal based on the profile: current plan, renewal history, estimated price sensitivity, and the competing operator's offer where available. The retention team opens the morning with the customer list ordered by switch probability, with the proposal ready for the call.
Senior reps on high-value customers, regulatory audit always traced.
The retention manager assigns the cases: high-probability, high-value customers go to the senior reps; medium-probability customers go through standard management. Decisions (proposal accepted, customer lost, renewal on a different offer) are recorded in the CRM. The trace of every retention intervention stays in the runtime audit registry for regulatory audit.
Declarative rules from the retention team, integrations during delivery.
Rules are declarative. The telco retention team defines in a human-readable format the switching patterns (expiry dates, consumption drop thresholds, portability signals), the personalised proposals per customer profile, and the team assignment rules. Integration with the telco CRM, the billing system, and the portability system is delivered during the project by the Exelab team. These systems vary significantly across operators; the adapter configuration is part of the delivery for the customer's case.
- 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
- Native notification channels
- Slack, Telegram, WhatsApp, OpenAI-compatible HTTP
- Telco CRM integration
- dedicated adapter during delivery (systems vary by operator)
- Billing and portability system integration
- dedicated adapter during delivery
- Scheduling
- configurable per instance (typical: nightly daily cycle)
- Memory
- persistent per instance
- Registry
- immutable, regulatory audit inspectable
Frequently asked questions about the agent.
Integration with the telco CRM, billing system, and portability system is delivered during the project by the Exelab team through dedicated adapters. These systems vary significantly across operators (proprietary systems, heterogeneous BSS/OSS). The adapter configuration is part of the delivery for the customer's case.
The typical pattern for Customer Win-Back is 8-12 weeks. Discovery 1-2 weeks, switching pattern rules and proposal configuration 2-3 weeks, integration with the CRM and portability systems 3-4 weeks, hand-off to the retention team 1-2 weeks. The main factor is the complexity of the telco CRM and the portability system.
The customer's telco retention team, via the graphical admin panel or directly in TypeScript. When a new commercial offer is introduced or the consumption-drop threshold for identifying at-risk customers is updated, the declarative rule is updated in the repository. For those who want to delegate operational management too, Exelab provides managed services: details are at /managed.
Value is measured on the retention rate of customers identified by the agent who, after the team's intervention, renew their contract — compared against the pre-agent baseline. The baseline is established during discovery. Customer Win-Back is one of the agents the customer can choose to build for their retention team.
From a 30-minute conversation to the squad in production.
A 30-45 minute conversation to understand how Customer Win-Back would be configured for the customer's case. Telco CRM systems, portability, switching patterns, proposal rules.