Quick and Effective AI-Enabled Automation

We’re living through one of the most transformational periods of growth in automation and AI that the world has ever seen. Goldman Sachs projects AI to grow worker productivity by 1.5 percentage points per year for the next ten years in the US. Leveraging these new technologies to help your business and deliver for your clients is key to success in an increasingly homogeneous, commoditized, and growing market. AI integrations are a dime a dozen already and when considering how to use AI for your business, we believe you should prioritize several key factors to ensure your implementation is useful and cost-effective: 

1. Build meaningfully unique tools that leverage your data and IP to provide unique outputs.
2. Prioritize flexibility and scale rather than task specific tools. 
3. Integrate tools and automation into your current workflows. 
4. Use AI for first pass analysis then go deep into the details with your experts. 
5. Don’t take AI for granted. Having AI products and processes is not better by default.
   • When deployed properly, AI integrations can increase both the quality and quantity of your work. 
   • When deployed poorly, AI is not only not helpful, but can be a significant detractor to your client relationships and overall output quality. 

AI is a lot more flexible than you might imagine. Don’t hesitate to push the boundaries of how it integrates with your workflows, especially if you have the experts on staff that can understand the outputs and build on them improve their work: 

• AI can understand tabular data and quickly identify trends, outliers, and insights that analysts might miss. 
• The structure of your tool is key to getting usable and helpful outputs from your process.  Ensure your tool balances input flexibility, user input, and output detail to maximize its useability. 
• Pair simple data manipulation tools with your AI integrations to maximize the amount of useful data that they analyze. 
• There are off the shelf solutions to format almost any document for AI ingest. 

At Mostly Human, we use AI tools for many aspects of our internal and client projects. We recommend to many of our clients to introduce these strategies to capitalize on their in-house IP and simplify workflows: 

• Use AI as an intelligent search engine for your documents. You can find project details, ask for examples, and even generate full documents and case studies with the right process. 
• Analyze results and identify opportunities for further analysis. This can be ad hoc using a purpose-built tool or, ideally, it’s part of your standard process. 
• Quickly create materials that match your style and follow your priorities. AI can create draft or final product work that matches your defined style and emphasizes priorities.   
• Automate repetitive reports and tasks whenever possible. Defined processes are easy to automate with minimal investment and can be paired with other tools to cut costs and improve the product. 

All sounds pretty good right? But you might (and probably should) be wondering: Is my data safe? It is… if you use custom integrations* and / or local models. This isn’t the case when using web apps like ChatGPT and is a major differentiator for you and your clients.   

*Note: Google may use data passed through its API to improve its models and should not be considered if privacy is a primary concern.