Digital Transformation: How NiCE Drives Sustainable AI and Data Centre Efficiency
Operations leaders are under pressure to scale AI without blowing up energy, cost, or risk. NiCE's 2024 ESG results show what that balance looks like in practice: steady efficiency gains, tighter governance, and a honest read of growing Scope 3 from data centres as AI demand rises.
Here's what matters for Ops: lower emissions intensity despite expansion, stricter vendor energy standards, practical cloud and coding efficiencies, and a governance structure that keeps decisions moving.
Scorecard: What moved the needle in 2024
- 11% year-on-year reduction in combined Scope 1 and 2 emissions intensity, driven by office optimisation and infrastructure upgrades in a hybrid work model.
- Absolute emissions rose 1.8% to 6,581 tCO₂e as the company expanded to 26 countries.
- Israeli data centre operations began migrating to cloud (completion slated for 2025); Scope 3 data centre emissions increased to 783 tCO₂e (from 408 in 2023).
- Primary data centre partners committed to 100% renewable energy as a contract condition.
- Pune office lifted renewable share to 30% and achieved IGBC Platinum Certification.
AI at scale, without waste
NiCE invested 14% of its US$2.735bn revenue into R&D, with a heavy focus on Gen AI and agentic capabilities across CXone Mpower and Actimize. These platforms support monitoring 5+ billion financial transactions daily and handling 30 million pieces of digital evidence. That compute adds up-and so does the energy.
To offset the load, the team applied energy-efficient coding, multi-tenant architecture, and automated shutdowns of dev resources during off-hours. These practices align with the AWS Well-Architected Sustainability Pillar and give Ops a clear playbook to copy.
Data centre strategy: growth with guardrails
The cloud move in Israel came with a short-term emissions bump (Scope 3) but clearer long-term levers: stricter supplier energy requirements and better utilisation. NiCE made 100% renewable sourcing table stakes for its two main data centre partners-baked into how contracts are awarded.
On facilities, NiCE converted two of five electrical units to renewables in Pune, bringing the site to 30% renewables and earning IGBC Platinum. For Ops, that signals a practical approach: incremental conversions, audited certifications, and visible progress employees can see. Learn more about IGBC standards here: Indian Green Building Council.
Governance that enforces momentum
- ESG Steering Committee (formed 2023) met four times in 2024, led by CFO Beth Gaspich-finance-backed oversight keeps trade-offs grounded in numbers.
- In 2025, the Board expanded the Internal Audit Committee to include ESG oversight (now the Internal Audit and ESG Committee).
- Materiality now includes Community Involvement and Giving, plus Third-Party Risk Management-important for vendor accountability.
- External validation improved: MSCI moved from AA to AAA; EcoVadis score rose eight points to 60 (bronze).
Targets for 2025-2027: what Ops can plan against
- Waste: +10% recycling rates at major sites.
- Mobility: +10% EV adoption in Pune and Israel.
- Infrastructure mix: +10% increase in cloud spend vs. hardware.
- Talent pipeline: maintain 75% employee usage of LinkedIn Learning; fill ≥33% of roles internally by 2025 and 35% by 2027.
- Water: monitoring coverage expanded to 55% of office floor area (from 37% in 2023); total use at 16,639 m³.
What Operations teams can borrow right now
- Set vendor energy rules in contracts: require 100% renewable sourcing for colocations and cloud regions where available; include quarterly energy and PUE reporting.
- Adopt a FinOps + GreenOps cadence: track unit costs and unit emissions together (per API call, per training run, per TB stored, per user session).
- Kill idle compute: enforce automated shutdowns for non-prod after-hours; right-size instances weekly; prefer multi-tenant services over single-tenant where compliance allows.
- Local wins add up: convert circuits to renewables site by site; target certifications (e.g., IGBC, LEED) with measurable milestones.
- Prioritise the biggest workloads: profile the top 10 services by energy use; refactor hot paths and storage tiers before anything else.
- Make AI demand predictable: queue training jobs to off-peak windows; use reserved capacity or savings plans tied to your actual usage curve.
- Expand measurement: include Scope 3 for data centre services with supplier attestations and third-party audits where possible.
Key KPIs to track each month
- Emissions intensity: tCO₂e per $ revenue and per compute hour.
- Utilisation: percent of instances above 60% CPU/GPU utilisation during peak.
- Idle waste: hours of unused compute and storage snapshots.
- Renewables: percent of energy from verified renewable sources (by site and supplier).
- Cloud vs. hardware mix: percent of spend and compute time in cloud.
- Training efficiency: kWh per training run; time-to-train vs. model accuracy targets.
The honest constraint
NiCE shows that you can lower intensity while absolute emissions rise during AI expansion. The hard part is bending Scope 3 back down as data centre demand grows. That's where contract language, workload efficiency, and disciplined scheduling do the real work.
As CEO Scott Russell puts it, the focus is "making a measurable and lasting impact" through ethical practices, inclusive growth, climate initiatives, and people-first innovation. For Ops, that translates to measurable rules, transparent dashboards, and steady iteration.
Optional resource for Ops upskilling
If you're building AI, cloud, and automation skills across your Ops team, see course paths by role here: Complete AI Training - Courses by Job.
Your membership also unlocks:
