You have an amazing idea for a company. Now it’s time to build the platform. Will you take a monolithic approach to building, like most enterprise-focused companies adopt? Or will you embrace change?
Lightbulb moment
Most enterprise applications are built like elephants: Giant databases, high CPU machines, an inside data center, blocking architecture, heavy contracts and more.
In short, they are big, slow and clunky for the sake of functionality.
But this is the opposite of where the enterprise needs to go today. Rather, they need speed, agility and adaptability; built for their environment without sacrificing power. These elephants should begin to look and think more like a streak of cheetahs.
First, determine what you want to achieve with your platform. What are the key components? What functionality would you like to offer as you build-out your roadmap? How will you scale? Think about it now so you don’t wind up with a stack of cards that could tumble if you’re not prepared.
In my case, I knew that if we wanted to build the transformative platform we envisioned, I had to change the way I looked at system architecture, leaning into my background in consumer applications and distributed computing.
How to take your own leap forward
If you consider what has happened in enterprise architecture over time, you see a lot of tools stepping on each other’s territory to try to make things simple.
Many data stores have become search engines and vice versa, but in reality they do a poor job of handling anything outside of their core competency. Trying to be everything in one comes at a cost; systems will not be super efficient or intuitive. If you really want to build a next-generation application, you have to rely on individual tools.
We built the platform I co-founded, Albert Invent, using highly scalable, low latency single table design NoSQL (dynamoDB), and streamed data to warehouse and search — and we adopted 10s of microservices to deliver and process. This was a move that would not have been possible even a few years ago.
But with consumer technologies becoming a commodity and managed services available through AWS, building and deploying enterprise architecture no longer has to look and act like an elephant.
You can get infrastructure as code with the click of a button and create a distributed architecture that makes sense for your business. This gives you access to the most modern tools to build the most modern applications your industry has seen.
To get it right, hire engineers who know how to build consumer applications, who have the skills in their DNA. They will be integral to building a new breed of enterprise applications, especially with goals like scalability and performance.
With the right people in place and a modular programming approach, you can start creating innovative, best-of-breed software components through APIs that are highly specific to a given problem or function.
For example, we have an inventory component, a worksheet component, an experiments component and a regulatory component that serve the various tasks that occur in a lab. Each was designed and coded using the tools that made the most sense for the needs of particular users.
They didn’t have to compromise, which enabled us to create something people would enjoy using.
Going API-first enables everything to connect and all of the components to talk to each other. Infrastructure and application building changes forever.
Not only do you not have to build from scratch, you have the ability to customize, and pick and choose what you want to accept. APIs are quick to integrate, they’re stable, and you can use any language or framework. Simply, APIs provide the ultimate in flexibility. You may feel like you’re giving up some control, but in the best possible way — you’re trading off months of development time and costs with very minimal, if any, downside.
Making it to the other side
If you are in a data-rich, highly dynamic industry and you want to make a profound impact on the lives of your customers, shed some pounds and add some spots. Don’t be the elephant.
Become a streak of cheetahs to serve the needs of your business and your customers.
Neelesh Vaikhary is the co-founder and CTO of Albert Invent, an end-to-end R&D data platform being used by thousands of scientists worldwide. He has more than 20 years of experience building early-stage startups and products, as well as scaling high-performing teams and building distributed SaaS and PaaS systems across a range of industries. Previously, he architected several breakthrough software solutions including Citrix GoTo Meeting and end-to-end secure chat software for Symphony. He also spent nearly a decade as an architect with Autodesk, and built a consumer startup, like.com (acquired by Google), that used computer vision and machine learning to simplify photo sharing and shopping.
Illustration: Li-Anne Dias