How Aily Labs built the first enterprise decision intelligence mobile app using AWS

Fortune 500 enterprises struggle with making data-driven decisions across silos, limiting their ability to harness the power of AI in a coherent and productive way across the enterprise. Traditional approaches rely on static reporting and manual analysis, leading to delayed decision-making and missed opportunities. "Companies have data in R&D, manufacturing, supply chain, and commercial functions, but everybody lives in their own separate silos. There is no layer which combines all the data," says Bianca Anghelina, CEO of Aily Labs.

This creates significant operational inefficiencies. Decision-making processes become sluggish. Companies lack the ability to run complex scenario analyses at scale, while limited visibility across enterprise functions leads to suboptimal resource allocation and increased enterprise risk. These challenges get worse as organizations grow, resulting in missed opportunities and reduced competitive advantage.

This overload of daily decisions creates what Aily calls the decision paradox — so many choices, data points, and cross-functional dependencies that large organizations often freeze. The result is decision latency — the gap between data and action widening beyond what modern operations can tolerate.

Recognizing these challenges, Aily Labs envisioned a transformation in how enterprises operate. The company delivers an integrated decision intelligence ecosystem that connects company-wide data through AWS cloud infrastructure to provide end-to-end solutions for forecasting, simulations, optimization, and insights.  Aily’s mobile experience embeds decision intelligence directly into business routines by proactively surfacing insights the moment an agent detects new risks or opportunities, turning the app into a personalized, always-on decision advisor where answers arrive in real time. By providing proactive recommendations through AI and scaling to support thousands of concurrent users, all while maintaining enterprise-grade security and compliance, Aily Labs revolutionizes business decision-making.

"I had a vision to finally develop an app which is able to support organizations to take decisions in real time," says Anghelina. "I am really on a mission to democratize AI and empower everyone with AI without the need to be technical experts."

Aily Labs built its comprehensive decision intelligence platform on AWS infrastructure. At its core, the platform uses Amazon S3 as a central data lake, storing hundreds of terabytes of decision histories, features, and all the files and data that result from the training process of a large language model, known as LLM model artifacts. To ensure responsive performance for thousands of users, the company deployed its services and near real-time inference capabilities on Amazon Elastic Kubernetes Service (EKS) running on EC2 m6a.8xlarge, m7i-flex.xlarge, and m6a.4xlarge instances, with built-in autoscaling capabilities.

The platform leverages Anthropic’s Sonnet and Haiko models on Amazon Bedrock to securely integrate Large Language Models (LLMs), powering natural language interfaces and AI agents. For example, Aily Labs uses Bedrock Prompt Caching with its models to leverage the cache memory to skip recomputation of inputs, allowing Bedrock to share in the compute savings and lower response latencies. For data warehousing and analytics, Aily Labs employs Amazon Relational Database Service (RDS) to automate time-consuming database administrative tasks as well as benefit from push-button scaling, automated provisioning, automated patching, and built-in high availability and disaster recovery (HADR) capabilities.

Aily Labs is also able to natively integrate with AWS AI and ML services such as Amazon Textract to read and process any type of document with no manual effort, as well as Amazon Polly, an AI voice generator, to deploy high-quality, natural-sounding human voices in dozens of languages. The company also relies on AWS for comprehensive security through services such as AWS CloudTrail for operational and risk auditing, governance, and compliance,  and AWS Key Management Service (KMS) for creating and controlling encryption keys to protect data.

AWS enabled Aily to scale from a handful of foundational services to thousands in a matter of days, supporting rapid tenant onboarding and continuous feature expansion.
With Bedrock, Aily’s AI brain can dynamically select the best model for each use case, giving enterprises multi-model flexibility without vendor lock-in.

The resulting solution comprises the Aily Agent platform with over 1,000 specialized AI agents orchestrated by the Super Agent, a mobile-first app providing instant access to insights and decision action, and the Aily Correlator^TM system connecting enterprise-wide data.  The company uses Anthropic models from Bedrock for their agents, which are built on a Pydantic AI agent framework, using Amazon EKS to build, run, and scale production-ready applications easily across any environment. They are also currently experimenting with the training of small models using Amazon SageMaker using high-performance Amazon EC2 G5 Instances, with a goal of integrating analytics and AI for unified access to their data. "Before AI and the Aily app, decisions were made in silos based on one to three scenarios compared to each other,” according to Anghelina, “With our agents, this is completely changing. Agents can run hundreds of thousands of 'what if' scenarios in real time and propose an optimized scenario." Depending on autonomy levels defined by the enterprise, Aily’s agents can even go beyond recommendations to execute decisions directly within transactional systems. This architecturally moves enterprises from decision support to decision automation — intelligence that not only guides the business but runs it.

The impact of Aily Labs' AWS-powered platform has been transformative for enterprise customers. According to Anghelina, “The platform delivers up to 99% predictive accuracy in business forecasting, while our pharmaceutical clients have experienced an 83% reduction in inventory risk.”

The financial impact has been equally impressive, with projected value of $685M through inventory optimization and expected incremental net sales of over $550M through resource reallocation. The platform now supports over 300 AI models and thousands of agents, processing requests in real-time for tens of thousands of users. Aily’s impact is not limited to forecasting and supply chain efficiency; the platform delivers measurable P&L outcomes such as margin improvement, sales uplift, and real-time resource allocation for CEOs and CXOs navigating constant volatility.

Aily democratizes AI across the entire organization — from C-suite to frontline teams — creating one shared language and one source of truth across functions. This cultural shift, enabled by accessible mobile workflows, strengthens cross-functional execution and accelerates enterprise-wide adoption.  "AI is transforming the way companies operate from a static reactive world into a proactive, dynamic world of permanent decision making," says Anghelina. "This is intelligence that doesn't just support decisions. It automates them."