Each one follows the same arc: context, problem, what I did, result, what I learned.
Grew analytics from a single analyst role into a multidisciplinary data science, AI and analytics team, and widened its remit from consumer to consumer plus business CVM.
Repositioned CVM analytics from reporting into an in-house AI enablement layer — models, decisioning, orchestration and local infrastructure — that serves the telco and its adjacent businesses with no licensing cost.
On-premise AI that reads customer chats for sentiment and buying intent, explains bills in plain language, and automates callbacks and ticket handling — keeping customer data in-house.
A subscriber-level social-network and behaviour feature table, plus a model-training feature store in build, so every model starts from richer, consistent signals — lifting the inactivity model's precision.
A competitor-intelligence platform with near real-time visibility of device, plan and travel-eSIM pricing, and an audit suite that checks every sale against 10+ business rules instead of a sample.
A next-best-offer pipeline was rejected by the cluster's memory admission control; a redesign around narrow ranking and staged joins brought it back under the pool limit.
A proof of concept pairing weighted business rules with an Isolation Forest to surface anomalous digital-fulfilment activity, delivered to reviewers in Power BI.
Designed the analytics and growth framework for an insurance partnership — B2C segmentation and a B2B model for embedded insurance — then turned it into a deployed lead-scoring pipeline and a growth engine in build.
Defined the scope for an interaction-analytics and agent-intelligence programme with an external vendor — data sources, prioritised use cases, and success metrics.
A connected set of retention analytics for postpaid — contract-locking analysis, churn tracking, telesales next-best-offer targeting, and ARPU outlier detection.
Led the client-side resolution of an integration failure between the campaign platform and a real-time decisioning engine that had stalled campaign development with the vendor.
Beyond the case studies: every use case my team is building on our in-house AI enablement layer, across commercial, customer care, adjacent businesses and core CVM.
Adoption
Near real-time price and plan benchmarking against the market for devices, mobile plans and travel eSIMs — saving products and devices teams 3+ hours a week of manual tracking.
Adoption
Automated auditing of every sale against 10+ business rules, so quality and fraud checks cover all transactions instead of a sample.
Development
AI models that turn customer chats into subscriber-level buying signals for devices, broadband, roaming, gaming and rewards.
Development
Extending market monitoring to stock availability and forecasting device demand.
Adoption
An AI assistant that breaks a bill down and explains the charges in plain language — fewer disputes and shorter interactions.
Adoption
Sentiment analytics across customer chats, surfacing trends and hotspots for earlier detection of dissatisfaction and sharper agent coaching.
Enhancement
An automated workflow that queues, prioritises and triggers callbacks for missed contacts.
Enhancement
Automated distribution and prioritisation of customer-care tickets.
Development
Further AI use cases to support front-line agents in real time.
Deployed
A propensity-scored lead pipeline that directs insurance outreach to the highest-potential customers.
Development
An intelligence layer — analytics, modelling, profiling and insight on partner data — orchestrating the full insurance customer journey for the global platform and brokers.
Development
Extending the growth engine to brokers in a further stc Gulf market.
Deployed
Segmentation on transaction behaviour, engagement, status and product adoption to target customer strategies.
Development
Account-level churn, customer-stickiness and cross-sell propensity models feeding personalised retention and cross-sell campaigns.
Deployed
Automated prediction and targeting of prepaid customers likely to go inactive, for early intervention and reactivation.
Enhancement
Predictive scoring of the postpaid base for earlier, proactive retention — now adding network, app-behaviour and financial data.
Development
A next-best-action engine routing each ticket to an offer, a resolution or an escalation.
Development
Live alerts on KPI drops and leadership slides generated automatically from live data.