The best way to start an AI project? Don’t think about the models

Tuesday, March 07, 2023

Welcome to TechCrunch+ Tuesday

Image Credits: David Merron Photography / Getty Images

Approximately one in three restaurants will go out of business in its first year. For construction companies, that figure rises to 53%.

But AI projects are the real heartbreakers: A Gartner study found that 85% are destined to fail “due to bias in data, algorithms or the teams responsible for managing them.”

Unfortunately, the profound fear of missing out means many organizations are jumping into AI projects with both feet even though they don’t fully appreciate the scope of work involved.

“The best way to ensure you are on the correct AI development path is to start your AI project without thinking about the models,” recommends Eran Shlomo, co-founder and CEO of Dataloop.

“Most of the data that the AI needs to perform at its best ability is not available to the development team," he writes. “This creates a ‘chicken or egg’ problem: Businesses need production data to deliver a functional model, but the model needs to exist in order to go to production.”

In a post aimed at non-technical managers and senior developers, he shares a framework for building a core team consisting of data scientists, domain experts and data engineers who can build a system that can learn from its mistakes iteratively.

Via collaboration, “the AI provides automation, speed and low costs” while the team steers “the AI to a correct result in a constantly changing environment.”

According to Shlomo, working along these lines generates a machine learning data flywheel, “essentially planning a learning system rather than an AI model that works properly at a single point in time.”

Thanks very much for reading,

Walter Thompson
Editorial Manager, TechCrunch+
@yourprotagonist

Despite the downturn, CVC gains traction in Brazil's startup ecosystem image

Image Credits: Alex Sholom / Getty Images

Brazil’s corporate venture scene is very much in its early days, but in the last few years, companies in sectors like mining, telecom and retail have been getting into the game.

“These CVCs should be structurally advantageous for Brazil's startup ecosystem, as it introduces a stable pool of medium-term dry powder that could reduce volatility,” says hedge fund investment analyst Matheus Tavares Dos Santos.

Gatik's Gautam Narang on the importance of knowing your customer

Image Credits: Bryce Durbin

As the toxic train derailment in East Palestine, OH, illustrates, our brittle legacy supply chains are long overdue for an overhaul.

Autonomous vehicle startup Gatik operates approximately 40 driverless heavy-duty semi-trailer trucks on routes up to 300 miles long, connecting distribution centers with smaller hubs.

Rebecca Bellan interviewed Gatik’s CEO and co-founder, Gautam Narang, to learn more about the company’s operations, investor expectations and how a shortage of human drivers is impacting growth.

“We have not done any free delivery ever,” he says. “We have been doing commercial deliveries since 2019, meaning every trip that we have made, we have been paid for."

If you have more than one business model, you don't have a business model image

Image Credits: masterSergeant / Getty Images

If your startup doesn’t have a well-defined business model — and a backup in case the first one fails — it’s unlikely to get funded.

Haje Jan Kamps defines it as “the full stack of how your company operates: How you deploy your resources (money and people) to create products and attract paying customers, and how you retain those customers.”

There's no need to stumble in the dark: Seed-stage founders can largely rely on LTV and CAC to shape strategy, since identifying “a repeatable business model” is Job One.

“The important thing is to narrow down the focus of your business model and how you're going to focus your attention during the sales cycle of your product,” he writes.

Creating remote work rituals that stick image

Image Credits: peepo / Getty Images (Image has been modified)

Remote teams have a lot of flexibility when it comes to when and how they work, but adding some structure can enhance productivity and transparency without sacrificing freedom.

“Ultimately, asynchronous work only serves you when you compartmentalize phases of work with your team,” says Stefanie Palomino, chief product officer and general manager at ROOM3D.

This post offers several tips that can help managers deploy active listening techniques that foster engagement, improve communication and, ideally, reduce the number of meetings that take place.

“The routines people create are negotiated over time, but it's something we've come to take for granted.”