Session Overview
Day 2 doubled down on “Stronger, Faster, Measurable” with two big product spotlights: Exact Forecasting Bot and Intraday Spike Bot. Plus live customer stories from ID.me and CVS Aetna. The message was to pair better plans (interval-level accuracy across channels) with faster reactions (detect, diagnose, and contain unexpected surges in minutes).
- Exact Forecasting Bot: Adds AI atop best-fit algorithms to model cross-channel impacts (voice, chat, email, back office). Targeting >90% daily/interval accuracy, early adopters claim to see +5–10% accuracy, a potential $2–$4M annual savings in a 1,000-agent center.
- Intraday Spike Bot: Auto-detects anomalies, analyzes conversation content vs. baselines (yesterday/last week), quantifies impact (e.g., $30k/hour), and proposes actions: IVA script updates, agent coaching nudges, and real-time knowledge posts.
- Customer evidence:
- ID.me: Live chat mix rose 6% → 53% (goal 75%); self-service containment moved from low-60s to mid-70s.
- CVS Aetna: 100% AQM coverage for coaching, mined ~800k monthly surveys; insights informed hiring cadence, tenure, and support tied to material business outcomes.
- ID.me: Live chat mix rose 6% → 53% (goal 75%); self-service containment moved from low-60s to mid-70s.
- Theme: Move beyond queue shuffling. Use unified data + agentic updates to address root causes and not just firefight symptoms.
Industry Perspective
Enterprises are exiting pilot mode and demanding time-to-value with proof. Two macro data points truly highlight the moment:
- ROI is becoming visible: In a SiliconANGLE/theCUBE Research survey, 46% of respondents report tangible ROI from AI deployments, reflecting growing confidence when projects are scoped to business outcomes.
- Budgets are following outcomes: Among 600 IT decision-makers building proprietary AI apps, 83% expect to increase AI spend in 2025, with most planning double-digit growth, despite flat overall IT budgets.
Still, complexity and skill gaps slow modernization, a theme flagged repeatedly for app/CX leaders. Packaging forecasting and intraday response as opinionated bots helps reduce that complexity while keeping hybrid/on-prem constraints intact; key for conservative, regulated shops.
Overall, Day 2’s combo of plan accurately, react instantly, measure continuously is the right operating model for contact centers living with volatile demand and strict governance.
Moving Forward
1) Treat forecasting accuracy as an operating contract
- Define what “>90% every interval” means per channel/interval, set minimum data quality, and publish explainability views (weightings, uncertainty bands). Tie gains to overtime avoidance, shrinkage, and schedule efficiency.
2) Stand up intraday governance before the next spike
- Pre-agree triggers (∆ vs. yesterday/last week/same time), auto-actions (pre-approved IVA snippets, KB posts, agent cues), and approval roles with audit trails. Measure time-to-contain as a first-class KPI.
3) Make value measurement turnkey
- Use out-of-the-box dashboards to attribute impact: containment → capacity, AHT/ACW → labor $, coaching → error/callback reduction, survey mining → revenue retention. Establish 30/60/90-day checkpoints.
4) Embrace hybrid without duplicating builds
- Provide reference patterns for where logic lives (on-prem vs. cloud) and how changes propagate. Use IVA Studio to ship small changes in days, validate, iterate.
Bottom line: Day 2 translates CX automation into a closed loop where you can predict demand precisely, respond to volatility in real time, and prove outcomes on the dashboard. For leaders navigating vendor noise and governance drag, this is the practical recipe to scale AI while keeping service levels, costs, and revenue moving the right way.
