CTO’s Guide: Delivering Tangible AI Value With Our Expanded AWS Partnership

AI remains at peak hype, yet promising use cases continually appear. As retail businesses increase AI investments and expand experimentation, a clear AI playbook is emerging. What’s also clear is that clean and consumable unified data is the key to AI success. Retailers equipped to use order data effectively within AI will be the, ahem, clear winners.

But, omnichannel success on the commerce platform side requires a comprehensive approach with deep, sustainable integrations and partnerships. This creates serendipitous opportunities for building new customer experiences across the many platforms that power e-commerce, vendors that serve merchandising, order management, search, marketing, and more. Consider the value of intelligent cross-pollination between Product Discovery and Merchandising vendors to build hyper-personalization. These opportunities are seldom achievable with siloed, individual vendor-based strategies that lock customers into a specific vendor’s direction.

At fabric, we’ve set out to revolutionize commerce, and now we’ve expanded our partnership with Amazon Web Services (AWS) to bring more of these AI-powered commerce, merchandising, and fulfillment innovations to modern omnichannel businesses. Let me explain why this is so exciting for fabric customers.

Where’s the value in AI?

In my last Tech Thursday, I explored why AI is on its way to delivering on the hype for the retail industry. Retail-focused LLMs, generative models, chatbots, and other advanced technologies are aimed at solving reach challenges for retail customers, organizations, and even individual roles.

There are plenty of other promising new use cases evolving in the market. With fabric’s AI-powered assistant and AWS, retail businesses will benefit from these AI-driven capabilities:

• Take action with insights and recommendations on how your products, inventory, and order fulfillment are performing. 
• Optimize and automate order orchestration, fulfillment, and inventory allocation. 
• Improve conversion by enriching product data across descriptions and attributes. 
• Streamline how you onboard your suppliers’ third-party purchasing data to grow your product offerings quickly.

The current landscape of commerce solutions requires retailers to assemble an end-to-end customer experience with point solutions from three dozen or so existing vendors. Even in today’s API-rich world, data is not shared across these siloed systems easily. Retailers then bear the burden of paying for yet another middleware solution to munge the disparate data or build several data stores with ETL and other intermediate data preparation steps – all of which require additional maintenance and support. 

Fortunately for fabric customers, from our inception, we’ve reimagined the commerce platform to overcome these hurdles, increase speeds and customer convenience, and more. We’re committed to standardization for our data and APIs. We continue to build data solutions that are easy to manage, import, and export in real-time and in bulk to support coming ML-driven use cases.

Now, we’re doing even more with this expanded AWS partnership.

Harnessing the power of AI to revolutionize commerce for omnichannel retail

Why are we doubling down on AI? Pockets of the e-commerce world already benefit greatly from existing AI solutions; this is no mystery. Yet, nearly half of retailers still need help turning data into growth. With the progress of technology, there is a promise of uplifting the entire e-commerce ecosystem through iterative yet pivotal improvements using AI. 

Leveraging AWS infrastructure to drive AI at scale in retail

We’re leveraging AWS’s power and infrastructure to revolutionize the way retailers operate, drive innovation, and make better business decisions using AI. While every other commerce platform seems to be piling on the generative AI for marketing’s sake, we’re looking at AI as a broad opportunity for retailers, and AWS lets us give our customers more options, flexibility, and power.

AWS-powered AI models enhancing retailers’ operations

We’ve utilized AWS to train, test, and deliver custom and pre-trained models that power fabric’s composable OMS and commerce platform. Through Amazon Bedrock, fabric is utilizing AWS large language models (LLMs) to build new AI features. Retailers can expect a new level of access to knowledge and control using our AI solutions.

Continuous iteration on advanced AI techniques for retail transformation

We’re focused on putting game-changing capabilities into retailers’ hands today. In the weeds of LLMs, the market is still experimenting with how to build models, whether to take a specialized or general approach, when and where to inject the human-in-the-loop with safer and more accurate reinforcement learning from human feedback (RLHF) or go with faster but less accurate reinforcement learning (RL) techniques.

We expect this evolution to continue with breakthroughs along the way, which will bring large step changes to customer behaviors, efficiency, and cost of operations.

With this expanded partnership with AWS, we’re able to deliver unprecedented AI-driven value to supply chain managers, merchandising managers, and customer service representatives with faster, easier, and more relevant data insights, recommendations, and actions.

Data is the key for every omnichannel retailer, and fabric

It’s obvious that data is what creates value and success for every business these days. It’s also shocking that we’re still saying things like it’s important to be data-driven, it’s bad to have data silos, and it’s good to have easy access to accurate data. 

But that’s the nature of data, technology, and business: things are constantly changing. Social commerce didn’t exist a decade ago. E-commerce barely existed 30 years ago. AI is only beginning to enter the consumer experience. Every time a new channel or technology pops up, you deploy a new solution that captures data in a different way and, unfortunately, creates a new data silo.

We understand those realities, but we also see that AI will be the solution to many related problems. Here are a few fundamental principles that drive our approach to data and AI:

1. Data must be standardized and centralized to be useful. It shouldn’t matter where your data resides or how it’s structured. Standardization brings together data from wherever onto a single platform where retailers can extract value easily.
2. Value is the ultimate goal. Too many solution providers are creating AI solutions and LLMs just for the marketing bump. A new solution to increase conversions through AI-powered price optimization is great, but what about clickstream, competitor matching, and other data?
3. Our customers must reap the benefits of that value. We’ll build value for the operator in service of the end customer. An improvement for a retailer that does not translate into some improvement for their customers will rarely improve sales or loyalty. We are committed to the growth and sustainability of our retailers, and this will always be achieved by improving value for their customers.  
4. Focus on creating AI primitives across the retail experience. AI primitives are the fundamental set of components – data, logical pieces, and the delivery methods – all working together to create an AI solution. We’ll identify opportunities spanning every aspect of merchandising, fulfillment, and customer support. We’ll also create incremental and widespread value that results in compounding improvements in a retailer’s business; this will dictate the best platform requirements. 

Why we’re excited about the power of AI to help retailers unlock more value

We’ve said it before, fabric is on a mission to “revolutionize commerce for everyone.” Rethinking data is foundational to our mantra. Data, within fabric, is highly cleansed and standardized, because we had the foresight to prepare for this new world of omnichannel commerce.

We remain focused on solving some of the harder challenges of LLM hallucinations while improving accuracy to create the most efficient, consistent, and explainable models for commerce. As such, we continue to see opportunities to rethink the entire retail data flow and decision-making loop – it’s why we started fabric and why these new capabilities get us so excited. It’s also why fabric customers can rest assured that our entire team is focused on continuing to revolutionize the omnichannel commerce experience.

Expect to see new experiences that improve the retail operator and end-customer experiences using language and other models. We will continue to launch incremental value using modern AI techniques while keeping an eye on disruptive and industry-changing solutions. And expect to see even more related announcements at NRF in January, 2025.

Ready to see AI in action? Join us for our upcoming webinar, “The Future of Retail: Thriving in the Age of AI and Connected Commerce,” on Thursday, September 26 at 10:00 AM PT. Hear from retail experts Jay Topper and Jenna Flateman Posner as they dive into how AI is transforming retail operations. Discover strategies for optimizing omnichannel experiences, real-time inventory management, and future-proofing your business for the AI-driven retail revolution. Don’t miss this opportunity—sign up now!

Umer Sadiq

CTO @ fabric. Previously software development @ Amazon and Williams Sonoma.