getty
Making technical operations more sustainable doesn’t necessarily require major investments, new platforms or attention-grabbing initiatives. Companies can often reduce their environmental impact by improving efficiency, limiting waste and getting more value from the technology and infrastructure they already use.
A more practical approach also makes sustainability easier to measure and maintain over time rather than treating it as a one-off campaign. Below, members of Forbes Technology Council share straightforward ways companies can make their technical operations more sustainable without relying on lofty claims or unnecessary new products.
Retire Redundant Processes And Applications
Build efficiency into everyday operations. Just as every resource on a space mission is carefully managed to maximize performance, organizations should regularly eliminate redundant processes, retire unused applications and optimize existing infrastructure before adding new technology. Sustainable operations are not driven by grand gestures but by intentional decisions that improve resilience, reduce waste and create lasting value. - Shelli Brunswick, SB Global LLC
Right-Size Cloud Resources
The most sustainable move isn’t buying a new tool—it’s reclaiming the capacity you’re already paying for. Most teams overprovision their cloud environments by 40% to 60% as a hedge against risk, which quietly wastes both money and energy. Continuously right-sizing allocated resources to actual usage cuts cost and carbon at the same time, with zero new software and no customer-facing disruption. It’s boring, it compounds and almost nobody does it consistently. - Volodymyr Lebediev, Triangu
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Track Resource Intensity By Workflow
Treat resource intensity as a production metric. Track compute, storage and data transfer per completed customer workflow, then give each service owner a reduction target alongside reliability and cost. Sustainability improves when teams see which architectural choices create avoidable load, not through another green claim or platform purchase. - Rishi Katdare, Amazon Web Services
Build Sustainability Into Infrastructure Planning
One practical step is ensuring the infrastructure behind technology is built sustainably. Every AI model, cloud service and digital tool relies on physical assets like data centers, power systems and water resources. Rather than relying on flashy claims, development companies can focus on efficient design, responsible resource use and long-term planning that reduces impacts while supporting growing digital demand. - Charley Daitch, Rowan Digital Infrastructure
Reduce Rework Across Technical Operations
One practical step is to reduce operational rework. Failed releases, avoidable incidents, duplicate fixes and poorly governed AI-generated code all consume compute, engineering time and business attention. Sustainable technical operations are not just about using fewer servers; they are about building systems that do not need constant correction after they reach production. - Todd Bernson, BSC Analytics
Eliminate ‘Zombie’ Workloads
Start by retiring “zombie” workloads. Every company has “hotel rooms” still air-conditioned for guests who checked out years ago. Audit unused servers, dashboards, jobs and data copies monthly. Sustainability gets real when ops teams treat waste like a bug, not a branding campaign. - Joel Frenette, TravelFun.ai
Audit Infrastructure For Efficiency Gains
Measure and optimize what you already run. Most companies have no clear baseline for their cloud or infrastructure carbon footprint. Start there: Audit compute usage, eliminate idle resources and shift batch workloads to off-peak hours. Sustainability gains from operational discipline are immediate, measurable and cost-saving. No new product required; just intentional engineering. - Ajit Prakash, ZENSAR TECHNOLOGIES LTD
Validate New Ideas With Customers
Stay close to the customer. Float any new ideas and concepts by those most likely to respond positively, and ensure it’s a painkiller, not a vitamin. With a few reference customers, go ahead and build the new service, product or enhancement. This way, you will develop only what is useful, thus staying away from flashy and shiny things. Exceptions may need to be made. - Krishna Govindarao, LightMetrics
Tie Team Incentives To Efficiency
Include infrastructure efficiency in team objectives and bonus criteria. Having a separate FinOps team monitor cloud spend has a limited impact if the rest of the IT teams are unaware and unconcerned about the resource usage of their activities. This issue becomes even more important with the rapid rise of AI token consumption by AI agents. - Stephane Donze, AODocs
Measure And Remove Operational Waste
One practical way: Instrument your systems for waste, not just uptime. Most operations have idle compute, oversized files and data nobody reads, all burning energy quietly. In streaming, I learned you can’t cut what you don’t measure. Track cost per output, then trim. The savings are real and repeatable and need no new product or press release. - Maksym Mashnytskyi, BMPS Technology
Make Efficiency A Core Engineering KPI
Track cost per inference and energy per transaction as quarterly engineering KPIs alongside latency and accuracy. Most teams optimize performance and treat efficiency as an afterthought. Once efficiency has a named owner and a tracked number, teams consolidate models, prune dead pipelines and right-size compute. You’ll see material savings within two quarters—no new product needed. - Dr. Chiranjiv Roy, C5i.ai
Optimize Compute, Storage And AI Use
Reduce waste while preserving the ability to experiment and innovate. Look for idle cloud workloads, AI-generated code running without necessary checks and balances, unnecessary SaaS and enterprise software maintained simply to keep up with the Joneses, and an unoptimized data strategy that doesn’t align with usage patterns or users’ storage and retrieval needs. Focus on the basics of managing technical operations—reduce compute, optimize storage and govern AI. - Harjeet Gulati, Millipixels Interactive LLP
Consolidate Data To Limit Sprawl
One practical way is to stop creating multiple copies of the same data for different workloads. Instead of running separate systems for AI, analytics, backup and archival, enterprises should look to consolidate onto a single governed data layer so the same dataset can serve more than one purpose. That reduces storage sprawl, data movement and operational overhead while also making the environment simpler to run and scale. - Garima Kapoor, MinIO
Build Efficiency Into Software Development
As a development company, one practical way to improve sustainability is to build efficiency into the software lifecycle. We focus on writing optimized code, reducing unnecessary cloud usage and retiring unused applications. It lowers infrastructure costs, reduces resource consumption and creates a more sustainable digital environment without relying on flashy sustainability claims. - Bankim Chandra, Dotsquares LLC
Pursue Continuous, Incremental Improvement
Adopt the 1% improvement philosophy by layering new intelligence onto legacy strengths. With agentic AI, your systems learn from deviations, and agents tweak future actions to improve results based on real-time feedback. You create a self-tuning system that continuously improves your operations without the need for new products. - Eugene Sayan, Softheon
Measure Efficiency, Not Activity
Most companies waste far more energy and resources on bloated processes, redundant systems and unnecessary compute than they realize. Sustainable operations start with eliminating waste. It is not as exciting as new technology, but it usually delivers the biggest impact. - Nick Damoulakis, Orases
Use AI To Strengthen Operational Monitoring
TechOps is becoming more sustainable with the help of AI. Proactive system alert benchmarks need to be set and refreshed continually based not just on user-reported failures but on performance. In my world, AI is an active working partner, continually monitoring user-based failures that were not caught by set benchmarks and improving them accordingly. - Anisha Manvatkar, NVIDIA
Watch For Functional Overlap
The first thing is to understand your existing products and get familiar with the latest features and capabilities that are available in the tools your teams already use. If you find a significant overlap in functionality, it often means tools are not being fully utilized or newer capabilities within existing products have been missed simply because nobody has kept the knowledge up to date. That awareness helps you prevent buying new licenses just for the sake of hype. - Yuri Gubin, DataArt
Maximize Existing Resources
You don’t need new gear to run cleaner. Most waste hides in systems that are underused or run longer than they should. Better visibility into the systems you already own, plus software and predictive maintenance, cuts energy and service trips without buying anything new. Get more out of what you have. - Andy Zosel, Diebold Nixdorf
Audit The Technology Stack Regularly
The most practical thing is to stop treating data infrastructure as a backlog item. Most unsustainability in technical operations comes from accumulated debt, with layers of tooling added to solve specific problems that are never revisited. To solve this, many companies decide on going with a completely new product, but what’s more effective is a periodic, honest audit of what you actually use versus what you pay for and patch. - Paul Deraval, NinjaCat
>> Home