Welcome to another edition of Growth Talks!

Today, we spotlight Marin Smiljanic, Co-Founder and CEO of Omnisearch. Marin isthe co-founder and CEO of Omnisearch, a startup developing the next generationof search technology. With a background as a software engineer at Amazon,Facebook, and SingleStore, Marin brings years of experience from bothfast-growing startups and Big Tech.
His work at Amazon, particularly on AWS and Alexa, inspired the creation of Omnisearch.Marin’s deep technical knowledge, rooted in his early passion for algorithmsand programming, now drives his vision for building a 'search for everything'platform that leverages his deep experience.
We sat down with him and asked a few questions regarding his startup journeyand the growth of Omnisearch.

What is the backstory of Omnisearch? How did you get froman idea to a business?

It was all fairly textbook. We built aquick prototype that was focusing on just audio and we ran with it for a while.We had the opportunity to partner with a company called Thinkific (kind of likea Shopify for online courses), and this led to our first paying customers. Wethen raised a bit of venture capital and started building the team.

What problem were you trying to solve when you foundedOmnisearch, and how has that evolved?

As with almost every startup, Omnisearchwas rooted in a problem I faced myself. During my time at Amazon, my colleaguesand I had an annoying problem: we had to consume lots of internal trainingvideos. They were very technical in nature, often whiteboard sessions where theold guard would explain system architecture and algorithms. They were alsopretty long, sometimes over an hour, so finding information inside them wastime-consuming. We wanted to solve that problem and package it into a product.
I think the main change since that time was that Omnisearch expanded from anaudio-only search tool to a proper “multimodal” one, covering video, images,and documents as well.

What was the biggest challenge you faced whentransitioning from a software engineer to a startup founder, and how did youovercome it?

Mostly the fact that I knew next to nothingabout sales and marketing. Granted, I did have a bit of training atSingleStore, since the company was still pretty small at the time, but the factis that those things don’t come naturally to engineers. I guess I adaptedpretty well.

From your perspective, what growth strategies have provenmost effective for Omnisearch in its early stages?

For the early stage, our partnership withThinkific was really important. Just being able to offer our software as asimple plugin on their platform was great. It gave us a seal of approval wewould not have had if we’d been selling solo. But we also recognized fairlyearly on that the money was in the enterprise, so we switched our go-to-marketstrategy accordingly.

Omnisearch focused on AI even when it wasn't that popularin the tech world. How has your perspective on AI evolved since then?

I think the technology has made some trulyamazing strides, especially with LLMs. Deep learning became thestate-of-the-art approach to machine learning back in 2012 with AlexNet, andit’s been accelerating ever since. I was particularly impressed by the resultsof the new OpenAI o1 models.

Are there any emerging trends, apart from AI, intechnology that you believe will significantly impact Omnisearch’s growth?

A big trend we’re seeing is that customers,especially larger ones, want more control over their data. This is why we’vearchitected Omnisearch in a self-contained way, so that it can be deployed bothas SaaS, on-premise, or in a private cloud.

How do you envision the future of video search technology,and where does Omnisearch fit into that landscape?

I think the main focus is going to beefficiency and performance. The perception of multimodal search solutions todayis that they’re very impressive in a demo setting, but not great at scale. Ibelieve we’re in a unique position to change that narrative.

Looking back at your journey, what do you wish you hadknown when you started your company?

The importance of sales. There’s no 'if youbuild it, they will come,' especially on something as cutting-edge as this.

What advice would you give to fellow founders looking togrow their companies?

Iterate as fast as you can, and make sureyour financial math works out. Long, enterprise sales processes should impactthe pricing, otherwise, you’ll never scale.

‍