Reworking enterprise operations: 4 high-impact use instances with Amazon Nova

Because the launch of Amazon Nova at AWS re:Invent 2024, we’ve got seen adoption tendencies throughout industries, with notable beneficial properties in operational effectivity, compliance, and buyer satisfaction. With its capabilities in safe, multimodal AI and area customization, Nova is enhancing workflows and enabling value efficiencies throughout core use instances.

On this put up, we share 4 high-impact, extensively adopted use instances constructed with Nova in Amazon Bedrock, supported by real-world prospects deployments, choices out there from AWS companions, and experiences. These examples are perfect for organizations researching their very own AI adoption methods and use instances throughout industries.

Customer support

Conventional chatbots usually frustrate customers with scripted, rigid responses that fail to grasp context or intent. For enterprises, these are missed alternatives to resolve points shortly, decrease help prices, and drive buyer loyalty. AI-powered purposes can perceive pure language, adapt to particular person buyer wants, and combine with backend techniques in actual time. Organizations are remodeling help from a value middle right into a strategic driver of satisfaction and retention. These are sometimes high-volume and interactive situations, so the steadiness of value, velocity, and intelligence is vital.

Customer support purposes constructed with Nova in Amazon Bedrock can seamlessly combine with enterprise information saved with AWS, and provide the safety, privateness, and reliability for manufacturing use in enterprise environments.

Infosys, a number one world IT companies and consulting group, developed Infosys Occasion AI for real-time transcription, multilingual translation, and clever summarization of reside occasion content material. Infosys Occasion AI is constructed with Amazon Nova Professional in Amazon Bedrock. Throughout a current occasion in Bangalore, the AI assistant dealt with round 230 customers per minute and was queried a median of 57 occasions per minute, producing greater than 9,000 session summaries. This answer enhanced information retention, engagement, and inclusivity by making occasion insights immediately accessible in a number of languages and codecs for hearing-impaired and distant individuals. By remodeling occasion content material right into a persistent, searchable multilingual information asset, Infosys Occasion AI accelerates studying and collaboration.
Fortinet, an AWS Accomplice and cybersecurity firm, makes use of Amazon Nova Micro to energy its AI help assistant, delivering important efficiency enhancements at a fraction of the price. By switching to Nova Micro in Amazon Bedrock, Fortinet achieved an 85 occasions discount in inference prices, dramatically decreasing TCO whereas sustaining speedy response occasions. The assistant now helps customers shortly navigate complicated documentation throughout greater than 60 merchandise, enhancing help effectivity and elevating buyer satisfaction.
Amazon Buyer Service makes use of Nova with its AI-driven problem decision system. The system is a two-step method combining intent detection and problem decision. Amazon Buyer Service custom-made Nova Micro, leading to 76.9% accuracy for in-domain points and 69.2% in generalization testing, surpassing present baselines by 5.4% and seven.3%, respectively. Moreover, Nova Lite is used for device choice, reaching 86.1% accuracy and 4.8% enchancment over current techniques.
AWS Summit New York Metropolis 2025 was attended by 18,000 individuals, that includes the AI assistant Diana for customer support developed with Nova Sonic. By dialing a cellphone quantity, the Sonic-powered voice assistant answered lots of of queries in regards to the occasion, together with session particulars, location, and FAQs.

Search

Giant enterprises face gradual, siloed, and inefficient search throughout huge shops of structured and unstructured information, costing time, productiveness, and buyer responsiveness. By adopting AI-powered, multimodal search that understands pure language and enforces safe entry, organizations can ship instantaneous, related solutions from paperwork, photos, and technical recordsdata. This accelerates decision-making, shortens deal cycles, improves buyer satisfaction, and reduces the price of information discovery at scale. Search purposes more and more depend on a mixture of data throughout modalities, together with textual content, paperwork, photos, and video.

Nova is among the many quickest and most cost-effective multimodal fashions, providing imaginative and prescient fine-tuning capabilities. Nova additionally integrates with broader Amazon fashions together with Amazon Titan Multimodal Embeddings and information companies together with Amazon OpenSearch Service for extra sturdy search capabilities and efficiency.

Siemens confronted rising efficiency bottlenecks as its huge datasets strained conventional search techniques, slowing retrieval speeds and impacting productiveness throughout its world operations. To handle this, Siemens built-in Amazon Nova, reaching a threefold increase in search efficiency that dramatically accelerated information retrieval and improved workflow effectivity. Amazon Nova delivers high-speed, scalable search capabilities, and Siemens’s implementation facilitates seamless integration with current techniques, sustaining enterprise continuity with minimal disruption. This enhanced consumer expertise and positioned Siemens to deal with future information development with ease, supported by steady efficiency monitoring and tight infrastructure alignment.
CBRE International Pulse System (GPS)—constructed with Amazon Nova Professional in Amazon Bedrock and OpenSearch Service—transforms property search throughout 1000’s of customers worldwide. In-built partnership with AWS Skilled Providers and GenAI Specialists, GPS replaces gradual, fragmented legacy techniques with an AI-driven, multimodal search platform able to dealing with complicated queries, huge PDFs, and strict permission controls. Key outcomes embrace 75% quicker doc ingestion, 70% decrease database latency, 87% quicker key phrase searches, and 51% quicker pure language queries. When absolutely deployed to over 6,000 customers later in 2025, GPS is projected to avoid wasting over 98,000 worker workdays yearly, unlocking $320,000 ARR and important operational effectivity. By shifting from Anthropic’s Claude Sonnet to Nova Professional and Anthropic’s Claude Haiku 3, CBRE additionally lower AI inference prices by 3.5 occasions and 12 occasions, respectively, with out sacrificing accuracy.

Video understanding and evaluation

Organizations are adopting video understanding purposes to drive enterprise worth throughout a number of fronts, together with buyer habits evaluation, site visitors patterns, and manufacturing high quality management. Safety and security advantages are realized via real-time risk detection and office security monitoring, and buyer expertise is enhanced via personalised content material suggestions and improved content material searchability. Organizations achieve aggressive benefit via data-driven decision-making and innovation in service supply, whereas lowering prices by minimizing handbook evaluation processes and reducing safety incidents. This complete method to video evaluation helps corporations extract insights from their video information, finally resulting in improved operations, higher decision-making, and enhanced buyer experiences. As builders construct, iterate, and evolve these purposes, there’s a rising demand to natively perceive video versus coping with the overhead of frames, time stamps, and synchronization.

Amazon Nova fashions can analyze, classify, and summarize data within the video primarily based on supplied directions. Functions constructed with Nova understanding fashions in Amazon Bedrock provide complete evaluation of a number of video codecs via versatile enter strategies, with the power to investigate, classify, and summarize video content material whereas dealing with recordsdata as much as 1 GB via Amazon Simple Storage Service (Amazon S3) integration.

Bitcentral partnered with Caylent to remodel how archived content material is found, accessed, and reused. Utilizing Nova Professional in Amazon Bedrock, Caylent deployed an answer that aligned with the wants of journalists, producers, and broadcasters throughout greater than 1,600 consumer websites. By embedding semantic video search, contextual metadata era, and AI-powered content material evaluation into its workflows, Bitcentral redefined how archived footage is listed, found, and reused. Journalists and producers can now floor high-value content material in actual time and unlock new income streams.
Loka, an AWS Premier Accomplice, constructed a video surveillance providing to routinely determine and classify tens of millions of visible occasions in video footage. This method successfully distinguishes between routine occasions and demanding incidents, serving to filter out non-essential actions and alerts. The answer proved extremely profitable, lowering irrelevant alerts by 55% whereas sustaining a risk detection fee above 97%. By implementing this automated filtering system, Loka doubled video monitoring effectivity for his or her consumer. The device, constructed on Amazon Bedrock utilizing Amazon Nova Professional, considerably decreased the workload for human operators whereas enhancing total risk detection capabilities.
Accenture Highlight can analyze long-form movies and routinely generate personalised short-form clips and highlights, that are notably helpful for sports activities content material like soccer, Method 1, and rugby. Highlight is able to matching content material to particular viewers demographics and may course of real-time CCTV footage in retail settings to create personalised presents. The system is constructed with Amazon Nova in Amazon Bedrock and operates via three specialised tremendous brokers working beneath a central orchestrator. Highlight can course of movies in minutes reasonably than the standard hours or days, whereas reaching value financial savings which can be 10 occasions higher than typical strategies. The answer is flexible sufficient for use throughout completely different industries, from media and leisure to retail, whereas sustaining top quality requirements and model alignment via its human-in-the-loop high quality assurance choice.

Inventive content material era

Organizations are searching for methods to revolutionize inventive content material era together with inventory imagery, advertising marketing campaign property, and product visualizations. It’s usually slowed down by fragmented workflows, excessive manufacturing prices, and the necessity to constantly steadiness scale with personalization. Advertising groups wrestle to maintain up with the demand for contemporary, high-quality property throughout a number of channels, whereas inventive fatigue and lengthy lead occasions restrict their agility.

Amazon Nova addresses these challenges with Nova Canvas and Nova Reel: high-quality inventive fashions that rework textual content and picture inputs into professional-grade photos and movies. Nova inventive fashions are designed to ship customizable visible content material with management options, making inventive content material era accessible and environment friendly for media, leisure, retail, advertising, and promoting industries.

Dentsu is reimagining how advertisements come to life with Amazon Nova inventive era fashions. What used to take weeks of brainstorming, filming, and modifying now occurs in days. Their inventive groups can sketch out an concept in plain language and watch it flip into polished movies and customized photos, prepared for markets throughout the globe in over 200 languages. Constructed-in safeguards like moderation, watermarking, and IP indemnity imply every bit stays model secure. For Co-op, Dentsu went a step additional—pairing Nova with Amazon Adverts to design customized viewers profiles that delivered a +4-point carry in model choice amongst 25–34-year-olds and a +5-point carry in favorability amongst prosperous buyers.
Quantiphi, an AWS Premier International Consulting Accomplice, developed Qreator, a generative AI-powered advertising content material creation service constructed on AWS. Their service helps entrepreneurs create content material via pure language prompts whereas sustaining model consistency and cross-channel adaptability. With Qreator, enterprise can obtain an approximate 30% discount in content material creation time and get to market roughly 40% quicker, automating what was a handbook course of, and enhancing consistency throughout codecs and channels.
The Fragrance Lab is a singular AWS activation that was showcased on the Cannes Lions Worldwide Pageant of Creativity. It demonstrates how you can construct personalised merchandise and marketing campaign property utilizing Amazon Nova basis fashions in Amazon Bedrock. Though our activation at Cannes Lions centered on personalised perfume growth and advert marketing campaign creation, the underlying structure and methodology may be tailored throughout various classes, equivalent to trend, meals, and beverage, opening limitless potentialities for custom-made buyer experiences. The Perfume Lab activation gained two International Business Awards: Gold for Exhibition Occasion Expertise and Silver for Experiential Occasion.

Conclusion

The 4 use instances offered on this put up reveal the utility of Amazon Nova throughout industries and purposes. From Infosys’s Occasion AI enhancing accessibility and engagement, to CBRE’s revolutionary property search system, to Loka’s clever video surveillance, and Dentsu’s inventive content material era, every implementation showcases important, measurable enhancements in effectivity, value discount, and buyer satisfaction.

Organizations utilizing Amazon Nova are reaching tangible enterprise outcomes via evidence-based adoption methods. By partnering with Amazon and AWS Companions, organizations are accelerating their AI transformation whereas sustaining sturdy foundations in safety, compliance, and privacy-by-design ideas.

To get began constructing with Nova, go to the Amazon Nova user guide or go to the AWS console.

Concerning the Authors

Abhinav Bhargava is a Sr Product Advertising Supervisor at AWS on the Amazon Nova group, the place he focuses on scaling generative AI adoption via customer-centric options. With a background in design and sustainability, he brings a singular perspective to connecting know-how and creativity to drive enterprise innovation. Based mostly in Seattle, Abhinav enjoys enjoying volleyball, touring, and studying about new cultures.

Raechel Frick is a Sr Product Advertising Supervisor at AWS. With over 20 years of expertise within the tech business, she brings a customer-first method and development mindset to constructing built-in advertising applications. Based mostly within the higher Seattle space, Raechel balances her skilled life with being a soccer mother and after-school carpool supervisor, demonstrating her means to excel each within the company world and household life.