ERPs are beyond most companies
Mention SAP and the image that comes to mind is probably one of a dominant force in enterprise software, known for its robust functionality and extensive suite of solutions. Or perhaps, it’s the image of IT consultants in shirts and ties configuring the tens of thousands of tables that comes with a basic installation. No matter the image, one cannot deny that SAP is both powerful and complex… and so are many enterprise resource planning (ERP) software.
The complexity (and cost) of these ERP software made it impossible for companies other than large corporations to adopt them. We are talking about small businesses or small-to-mid tech startups. So what happens when such a company needs to manage operational data or business processes?
Limited options for smaller organizations
If you’re a small business without your own tech team, you have a couple of options: you can either hire a developer to build an application for you, or buy an application off-the-shelve. Such applications are typically specific for a use case, for example, an application that manages inventory or an application for managing human resources. These off-the-shelve applications also require an admin to configure, such as adding your own data, though they are much less complicated than configuring a full-fledged ERP system.
If you’re a tech startup, apart from the usual options available to small businesses, you may also build your own internal applications. Over the years, some startups have made building internal applications much easier (we are huge fans of Retool). Despite these advancements, this option still requires a team of developers and man-hours, not just to develop the applications, but also to maintain and run them. The process of building these applications are also highly technical — involving dragging and dropping UI components and hooking up APIs or SQL statements.
Introducing Onelook
Here at Onelook, we want to make it easy for anyone to build their own ERP regardless of the business requirements. You could be a small school that is building something simple to track device loan requests. You could be a business owner tracking inventory. Onelook lets you build them using plain human language. We are leveraging the power of AI to do this.
Our principles
To make this possible, we follow these design principles when building Onelook:
Start from the business logic
In Onelook, users describe what their end goal. We use AI to determine what data is required for the application and create the schema for it. Think of it as a human engineer translating business requirements into database tables. Using the business logic, the AI also determines the functionalities for the application.
Meticulously guided, not opened and lost
There is no single method for generating a working application that is guaranteed to work, but there is one that is guaranteed to fail — dump all the instructions in one message and send it to the AI model.
Onelook is carefully designed so that users are not compelled to relay all the information at once. Too much information overwhelms not just AI, but also the human. Onelook encourages users to use a top-down approach and the whole process of creating an application feels like a chat with a human. Users first start with a simple description of the application they need. Onelook leads the conversation and asks for more details as the conversation proceeds. At each critical step, Onelook describes its understanding and what it has come up with to seek confirmation with the user.
Clarify any ambiguity
Using an application that is generated by AI in daily business operations sounds scary. And it sounds scary because it feels like the process in which the application was created causes the creator to have no control over the end result. Engineering (writing code) is assuring because it is precise. Sure, there would be bugs but these can be resolved by an engineer investigating it, fixing it, and testing the fix.
To mitigate this, Onelook will always ask users for confirmation when there is any ambiguity, especially when it comes to modifying data. For example, if the user asked for a feature to update data but did not make it clear whether it is supposed to be a batch update, Onelook will ask the user for more details.
Features
Human language to application
This is the core of what Onelook is. Chat with Onelook like a user talking to a product manager and a working application will be generated for you.
Multi-views
Sometimes the business process requires input from non-privileged users. Using our loan request application as an example, you would need a public view for your organization members to submit a loan request. You also need an admin view to manage these requests. Onelook allows you to generate these views.
Analytics included
Onelook comes with built-in analytics. View your data, group them, sort them. Forget about exporting your data to a separate data warehouse or connecting it to a dashboard application.
Your application with instruction manual
Applications may not be straightforward to use. It may not be obvious what certain fields are for and what values should be entered. Onelook-generated applications come with documentation that guides the end users. We use the business logic as a starting point to create these documentations and users can update them if they want. You never have to write a separate documentation from scratch or store them separately. With Onelook, your app and documentation are in one place.
More powerful than forms, less complex than ERPs
“If you can imagine it, you can create it.”
— I first heard of this from a professor during a software engineering module in college. This quote (a misquote?) has always stuck with me ever since. This is our vision and what we truly believe in Onelook. With the power of AI, this statement has never been more true.
Start using Onelook and unlock possibilities.
Try our demo or request for early access.
