Perplexity Partners with American Express: Ready-to-Use AI Skill Library for Small Business Owners

Perplexity has announced a curated library of ready-to-use AI skills — selected by American Express — available to eligible Amex small business cardmembers through the Perplexity Computer platform. These skills are essentially pre-built workflows designed to handle everyday business tasks. From cash flow forecasting to marketing campaign generation, users simply select a task, fill in their business details, and let the AI do the work — no prompt writing required.

The "One-Person Team" Problem for Small Business Owners

Small business owners often wear every hat at once, doing the work of an entire team on their own: managing finances, running forecasts, handling marketing, and keeping daily operations afloat. Perplexity's announcement put it plainly — "small business owners are already doing the work of an entire team" — a line that cuts right to the heart of this audience's core challenge.

For small businesses without dedicated analysts or marketing staff, the barrier to using AI tools has always been a real obstacle. Most general-purpose AI assistants require users to write clear, well-structured prompts — an unfamiliar workflow for business owners who aren't versed in how AI systems think. That's precisely where this partnership aims to make a difference: by removing the burden of knowing "how to ask AI the right questions."

Pre-Built Workflows: Lowering the AI Adoption Barrier

The core design philosophy of this skill library is "pick it up and use it." Instead of writing instructions from scratch, cardmembers open Perplexity Computer, select a task — such as "forecast cash flow" or "generate a marketing campaign" — enter details relevant to their business, and the AI delivers results through a pre-set workflow.

This approach aligns with a broader industry trend toward "agent workflows" and "skill templates." Compared to open-ended conversation, pre-built workflows offer more predictable and consistent outputs. Workflows curated by a financial institution like American Express carry an added implication: the underlying task logic has been specifically tailored for small business scenarios, rather than offering generic, one-size-fits-all answers.

For tasks like cash flow forecasting — which are highly dependent on sound financial frameworks — a purpose-built skill template is especially valuable. It ensures the analytical structure fits how small businesses actually manage their finances, rather than leaving users to guess which data points to provide.

---

"Agent workflows" are one of the defining paradigms in AI applications today. They refer to AI systems that autonomously execute a series of steps to complete complex tasks, rather than simply answering a single question. Unlike traditional chat-based AI, agent workflows can decompose goals, invoke tools, process intermediate results, and deliver structured outputs. Perplexity Computer is Perplexity's computer-use AI platform — capable of executing tasks at the operating system level, such as opening applications, filling out forms, and scraping data. This allows pre-built skills to automate end-to-end workflows across multiple steps, rather than just generating a text-based recommendation. Combining agent capabilities with vertical-specific preset templates is the technical foundation behind this partnership's attempt to solve the "general-purpose AI isn't good enough" problem.

---

Where Financial Institutions Meet AI Platforms

American Express's decision to partner with Perplexity reflects a broader move by financial services firms to embed AI capabilities as value-added benefits within their cardholder ecosystems. For Amex, this is a differentiated retention strategy: small business cardmembers don't just get payment and credit services — they also get free access to AI tools customized for their operating context.

From a wider industry perspective, this kind of partnership points to a pragmatic path for AI deployment. Rather than expecting every user to become a prompt engineer, institutions with deep vertical knowledge — like Amex's understanding of small business finance — partner with AI platforms to distill that expertise into standardized, reusable skills. This "domain knowledge + AI platform" combination may be more effective at solving real industry problems than a general-purpose model alone.

---

"Prompt engineering" refers to the practice of carefully crafting input instructions to guide large language models toward high-quality outputs. In general-purpose AI assistant contexts, the quality of the prompt directly determines the usefulness of the output — meaning everyday users unfamiliar with AI mechanics often struggle to get the most out of these models. Encoding domain knowledge into pre-built workflows essentially transfers the complexity of prompt engineering to institutions with professional expertise. In this setup, American Express plays the role of an "invisible prompt engineer" — its years of accumulated knowledge about small business finance is baked into the skill templates, so end users only need to supply their business data without needing to understand how to frame questions for AI. This is also the core value proposition of the "Skills-as-a-Service" business model.

---

Open Questions Worth Watching

Publicly available details remain limited. This launch is restricted to eligible American Express small business cardmembers, and the specific eligibility requirements, the full catalog of skills, and the real-world output quality of these workflows have yet to be independently verified. Cash flow forecasting and marketing campaign generation are only two examples mentioned in the announcement — the true breadth and depth of the skill library will need to be tested through actual use.

Additionally, AI workflows that handle financial data raise important questions around data privacy and security. When small businesses feed their operational data into an AI system, the mechanisms that protect that data become a critical concern — one that any product in this space must answer clearly once it launches at scale.

For small business owners in the United States, this service offers a low-barrier entry point into practical AI tooling. For the broader industry, it's a case worth tracking to see whether the "Skills-as-a-Service" model can achieve meaningful scale.