Afresh: Eliminating Food Waste via Climate Tech

This post originally appeared on LinkedIn and is the second in a two-part series; if you haven’t read Matt’s post yet, be sure to give it a read as well.

When I met Matt in the weeks leading up to starting business school at Stanford, I was blown away by his passion for our food system. He couldn’t stop talking about food: theories of nutrition, the history of how we’re fed, GMOs, CRISPR, The Dorito Effect. He was relentless; he was both charming in his earnestness but also at times just a bit annoying. He was so convicted in his career arc to transform the food system for the betterment of human and environmental health.

I wasn’t wired like that. I didn’t have a laser-focus on a specific industry. I didn’t have a grand vision for the future. Heck, for most of my time in my undergrad and graduate engineering programs, I didn’t even know what type of engineer I wanted to be, eventually settling on “Mechatronics” —the ultimate “jack-of-all-trades, master-of-none” combination of software, electrical, and mechanical engineering.

I did, however, have two principles that I was sure about:

1) I wanted to build innovative and impactful technology

2) I wanted to work on climate change

And my career arc—while zigging and zagging—has adhered to those principles, from my first ever internship at what is now the Woods Institute, to the first robot I worked on in industry (an autonomous ocean robot used by NOAA’s Pacific Marine Environmental Laboratory to monitor ocean acidification).

So six years ago, as Matt and I discussed the foundational thesis that would beget Afresh—that building innovative, purpose-built tools for the fresh food supply chain could dramatically increase efficiency and thus eliminate food waste—I needed to determine whether such an endeavor would meet those two guiding principles.

Food waste and climate change

While food waste, particularly in the climate community, is a familiar and established driver of climate change, six years ago I had no idea about the extent of that impact and its relative contribution. If I wanted to work in climate, could I make orders of magnitude more impact by building electric cars?

What I learned was that our food system consumes a MASSIVE percentage of our planet’s resources. As Matt noted in his post, growing, producing, and distributing food consumes 30% of global energy use and 38% of total land use. Devastatingly, approximately ⅓ of all this food ends up as waste, which in turn has myriad impacts on climate change in particular and the environment in general. The inefficiency of this land use drives deforestation. The food that ends up in the dump creates huge methane emissions. A massive amount of clean water is wasted. The list goes on and on and has been empirically and quantifiably documented.

In short, the critical takeaway, as I dove in, was that the aggregate impact of food waste on climate change is enormous.

In fact, the impact of food waste on climate change is so immense that, according to Project Drawdown in its macroscopic study of which societal level efforts would most mitigate climate change, reducing Food Waste ranked 1st and 4th according to their two methodologies (out of > 90 initiatives)! Once I wrapped my head around this staggering information, I quickly checked off one of my two requirements.

Innovative and impactful technology

Given that reducing food waste emphatically met my desire to work on climate, next, I wanted to understand if we’d be building technology that I found intrinsically interesting and satisfying. Different folks are drawn to different technological challenges. Some people love pushing the bounds of knowledge in research labs (I tried that in an engineering lab in grad school; it wasn’t for me). Other people want to take moonshots and put the internet on Mars.

I like applying cutting-edge technology in a pragmatic, solution-oriented way, to solve tangible problems that make a substantial impact. And as we started to explore the fresh food supply chain, and how we could optimize its efficiency, it quickly became clear that it was going to be a super fun challenge.

One of the best opportunities for impact is an area under-penetrated by technology. I had a Fluid Mechanics professor who frequently said, “Don’t go make the rivet on a Boeing 737 1/10th of 1/10th of 1/10th of 1% more aerodynamic. Go find an area where fluid mechanics hasn’t been applied and make the 80/20 initial improvement.”

When we started to shadow in grocery stores, we discovered that green-field opportunity my professor had described. Despite troves of data on consumer buying patterns, grocers big and small were running their fresh departments on human intuition and frequently literal pencil and paper systems like you see below. This resulted in systemic over-ordering, which in turn drove food waste. If we could effectively inject technology, we could make that 80/20 leap forward.

But as we examined the space, we came to understand that the problem wasn’t trivial. As Matt covered in his post, the fresh food supply chain hasn’t benefited from technological innovation nearly as much as has the supply chain for processed foods. While that was partially due to extrinsic factors such as historical market dynamics, consumer trends, etc., it turned out that it also was because, heretofore, the technology to solve the problems in the far more complicated fresh supply chain didn’t exist.

In particular, as we investigated applying data science and forecasting to improve decision making, we quickly realized that more traditional methods wouldn’t work, due to unreliable data and the multitude of other fresh-specific factors that drove uncertainty.  With brittle, legacy systems, garbage in meant garbage out. Instead, we needed to turn to emerging disciplines that solved for decision-making in uncertain environments—techniques like reinforcement learning, which was gaining popularity in the world of autonomous vehicles and other applications. (To be clear, our co-founding CTO, Volo, deserves all the credit here.) We needed to take a fresh approach (I couldn’t help myself).

So putting it all together, as I sized up the holistic potential of starting Afresh, this was a chance to work on one of the single most impactful climate change initiatives, which also offered the possibility of zero-to-one technological innovation, and which would require the practical application of brand-new, emerging technologies. It was a profoundly exciting opportunity; and, spoiler alert, we went for it.

Six years later, it’s gone far better than we could have ever hoped (and also has been much more challenging than we ever could have imagined). The technological and operational problems we are tackling to build an end-to-end solution are complex and tangible. Beyond next generation algorithmic approaches to forecasting and decision-making, we’ve had to build incredibly intuitive user experiences to accommodate the grocery industry’s high-turnover, dynamic workforce; modern, scalable infrastructure to handle large quantities of challenging data; and operations and support teams to deftly handle integrations and on-going operations with customers with hundreds of thousands of employees. Every time we execute—by improving an algorithm, launching a new store, training a new user—we quantifiably increase our impact. It’s been a wild six years, and, with our Series B, it’s only accelerating.

So if you’re like Matt, and you started your own snack food brand, read every book on food, and listen to two food systems podcasts simultaneously (one in each ear), say Hi! Or if you’re simply like me and see the planet cooking, and have decided you want to work in Climate Tech, Afresh is a great place, too!

Specific motivation aside, we’re a group of folks who are uniform in our passion for the social impact of our work. And getting to work with such a like-valued group is one of, if not the biggest, joys of working here.

It’s a third guiding principle I’m adding to my checklist from here on out.