Creating AI Solutions For The Agriculture Industry With Naeem Zafar
Farmers or business owners should monitor stored grain because it greatly impacts the whole enterprise. That’s where artificial intelligence in agriculture comes into play. Chad Burmeister sits down with Naeem Zafar for a conversation about the role of AI in enterprise sales. Naeem Zafar is the CEO of TeleSense, a company that offers solutions for problems in specific industries that have been untouched by technology. Agriculture is on top of that list, so Naeem is deeply motivated to change the world in this industry. In this episode, he discusses how he crafted the solutions to the problems in agriculture, the importance of AI in productivity and how he transforms his passion for engineering into leadership.
Listen to the podcast here
Creating AI Solutions For The Agriculture Industry With Naeem Zafar
I'm here with Naeem Zafar, the Founder and CEO of a cool company called TeleSense. We've never talked AI in agriculture, so we're going to spend a little bit of time on that. We're also going to talk about Naeem's view on the world in enterprise sales and how AI will impact enterprise sales. Welcome to the show.
Thank you, Chad.
I'm familiar with the Bay Area. In fact, I did a run in San Jose, the Rock 'n' Roll Half Marathon years ago. I love your area in Cupertino and San Jose. With many years in the business, tell us a little bit about your role. What do you oversee? What are companies using TeleSense for?
I'm an electrical engineer. I had started many companies. I've given and sold my last company to Oracle that was in the mobile security space in 2013. I was looking at what's next, what's big and I came to the clarity that IoT, the Internet of Things, is going to play a huge role, wireless sensors and connecting devices, making sense out of it. You put a layer of artificial intelligence on top of it with machine learning and you begin to extract insights that would normally be missed. There are two industries which are perhaps least touched by technology. Agriculture is probably on top of that list. Construction could be another one. This is where we focused. What could we do in agriculture which can make a difference? We did a couple in three years looking at different areas.
The bottom line is post-harvest grain. Grain, once you harvest it, never improves in quality. It goes downhill. You harvest everything in October and in March, April, May, you got to store it because you're going to use it throughout the year. Grain has to be stored. Can you predict when it will go bad? When will the biological activity happen? When can the spoilers occur? Can you predict, “Hang on to this bean. Don't sell this one. The quality is going to hold you well for five months but number 2 and 3, get rid of them within two months?” It's all about profit, preservation, saving the world and putting an extra couple of bucks in the farmer's pocket.
I toured the Breckenridge Brewery in Colorado years ago on my brother's birthday. They had rolled out an AI to manage the mix. My nephew goes to Colorado School of Mines and so does my son in electrical engineering. My nephew was asking questions. He's like, "How's the AI doing when it comes to the mix?" They go, "For the first 2 to 3 weeks, it's terrible. It got it all wrong. We had to throw away a lot of batches but now that the computer model is trained, it does amazingly well." It was this little computer with an ethernet cable plugged in. I was like, "You mean that replaced a whole bunch of people who made those decisions in the past?" It was just within 0.00. It was perfect.
Focus on making a huge difference in an industry that isn’t given much attention to.
AI is going to play a huge role in every industry. It will be pervasive like the internet was. In '95, '96, companies were saying, "Do we need a website? Is it a fad?" It tests everything how we operate and do business. AI will have the same impact except bigger.
I asked someone who's into machine learning and big data. He's been at it for years. I said, "What do you think? Is it exponentially bigger or is it just about the same 20%? What is it?" He said 5 to 10 times bigger. That makes sense to me.
We already see that. The companies like Amazon are using data so well that dynamically changed the price of the thing based on who you are, which ZIP code you're calling from, which IP address and what time of the day. There's so much you can decipher and their suggestion is based on what you bought. "You may want to buy this book." "I do."
With airlines, if you look at a ticket that's $180, ten minutes later, it's $236. So that our audience can get to know you, a lot of times, I like to ask the question, “When you were younger, where were you raised? What was your passion when you were a child?”
I was born in Pakistan. I was there until high school and then I came to the US to study engineering. I always wanted to be an engineer. Even when I was ten years old, my favorite toy is Meccano, trying to construct models and create things. To me, an engineer is a great job because it teaches you how to break down problems into chewable chunks and put them back together. When you learn that skill, you can apply that to anything in life. It's all about problem-solving.
I could see it in my son when we play a card game. When you play the basic level, it's simple math. It's either I'm plus three and you're minus three based on the card you are playing. You then get into advanced algorithms. There are expansion packs with this game we're playing. His mind is working through all the Math problems. Me and my wife are just sitting back going, "I used to be pretty good at Math but not to the level of a nineteen-year-old engineering student." It is nice to see the brain working. That led you into engineering with a few successful companies. Talking to our audience, there's got to be some people who are looking to go into entrepreneurship. Can you talk a little bit about how you transform the passion of seeing the patterns and doing engineering to leading teams? What caused you to become a leader?
First of all, going to entrepreneurship, it's not a job. You pick to do that. It's a calling and not everybody gets the calling. Not everybody is even suited for the calling. It's a tough life. Most of the time, you get excited but most of the time, it's in a trough of sorrow and things that are kind of working but are not really working. You do it because you're fit for that, nothing else. I call them the misfits of society. They can’t hold a job. They piss off their bosses. They are the little oddballs. Those people that make good entrepreneurs.
I'll raise my hand on that one. You can ask most of my bosses.
You have to ask yourself, "Am I really an entrepreneur?" You can have a good life. You can have a good family life and comfortable living if you play your cards right. Entrepreneurship is hard and you do it because there’s nothing else. You almost can't sleep at night because of an idea that you have or something is bothering you so much you've got to do it. The fun part and why we do that. I lived most of my life in lots of troughs of sorrow because you put the problem you're going to work on. You pick the people that you want to work with. You make the rules.
AI In Agriculture: Engineering is a great job because it teaches you how to break down problems into chewable chunks and put them back together. When you learn that skill, you can apply that to anything in life.
If you want every Wednesday off, you can have every Wednesday off. This is a choice, the freedom it gives you. You can throw me into the Gobi Desert and I'm confident I'll be selling something to the local within three months. I have no idea what it is but I'll figure it out. That confidence is very positive for some people. You got to be careful before you choose it. It's a lifestyle choice. You don't think about vacations that normal people do. On your wedding anniversary, you're checking your email under the table, pissing everybody off. That's how it goes.
It's funny because I've also tried to hire sales teams. I've been a sales leader with 100 people on the team. If I have a customer, it's a 43% close rate prospect. If my CRO has a meeting, it's 35%. That's pretty darn close. That's a rounding error in my opinion but if you go to the sales reps we've had over the last years, it’s 11% to 17%. That's unacceptable. I can't scale to that. That's the final gap and chink in my armor that I need to figure out as a leader.
That is a hard thing to do. The good thing is AI is to the rescue because this is the area where AI is going to play a huge role. You can look at sentiment analysis. You can bring it to analyze the email conversation between your sales rep and your customers. Smart algorithms, they can figure out. Is your salesperson being true to themselves? Are they overly optimistic? Do they sandbag? With sentiment analysis, analyzing the conversation, the words your customers are using, you can have a layer on top of your sales forecast, a truth serum layer. Those products are not quite out there yet but they're getting close.
I met this SVP of Sales from Clari a couple of years ago. They've done some pretty good work with Zoom video and some others. There was another customer too. I know the VP will. That's evolving but you're right. It didn't traditionally see into the email but bringing all of that to your point, the Internet of Things, it's all accessible. I wonder if that's what's going on in the world with trackers on your phone and the rollouts. We all become endpoints and we already are. If you have Google on your browser, half of me thinks that by putting myself out there as an endpoint, you can cure a lot of health problems. You can preempt things and yet there's the privacy concern of, "Am I giving up too much information?" What are your thoughts on the AI ethical side of the equation? That's always an interesting part of the dialogue.
I don't have the answer but my thoughts are we all respect privacy and privacy should be a prime concern but we may be taking a simplistic approach to privacy. I don't want to see baby diaper ads because I'm in the stage of my life that it’s not relevant to me. The more you can customize what ads I see and what things come my way as an offer, so I don't have to go through a lot of unnecessary clutter, it is a plus column. On the negative column is that if my medical history becomes public to my employer, they may be reluctant to hire me. It needs some safeguards. That's another extreme.
There's a lot of gray areas in between. My overall short answer to your question is privacy should be taken seriously. This already started in Europe. We should be able to control and monetize our data. We are the end product for Facebook and companies like that but we get nothing in return. I see a future when I'll be able to set some dials. I can check $146 a month because advertisers use my data to target something, then I don't feel so bad. I'm getting something for me. I'm the product.
It becomes a win-win-win versus win-lose because you're the customer. The win is starting to become lower and lower.
I'm the product. I'm being sold except I'm getting nothing for it but there are changes. Companies are working on that. There could be a smarter model of how this can all become more palatable.
Entrepreneurship is not for everybody. It’s a tough life and a calling that not everybody's suited for.
Let's go tangential here because we'll get into agriculture in a minute. What about the news? I talked to a guy years ago who said, "If we could crowdsource a thing at the bottom that said, these are known people that know the information. If you do 500 thumbs up by all these known people and then there are only ten thumbs down, then you can say, 'That's real news.'" There's so much misinformation and even disinformation. I'm curious about your engineering brain. Have you thought through that dilemma? How would you think about solving that?
We have started a class project on that in my Berkeley class. It's similar to what you said. What I was proposing was that we have a good housekeeping seal equivalent on it or think of that like your credit score is a number but it's a complex algorithm. Imagine a number of 1 to 100. With every story, you can go to a site, which can show what is the truth serum number? I can say, "I only show the things 88% and above. Anything below 88%, I don't trust." Somebody will look for anything over 55%. Somebody said, "I need to see 98%." It's some kind of a rating, which is independently done. A credit score is independent. They are not always accurate but are a close indicator. There's a need for something like that, which can validate any posts with some third parties like Good Housekeeping or used to be that. Consumer Reports gives an independent number to each appliance before you purchase it.
Some SVP at Riverbed years ago said, "Chad, you've got 100 salespeople on your team, but they're not quota-carrying. They drive for meetings, SQLs, MQLs and that whole thing.” He said, “What’s the one number that you can share with me that if they’re a federal salesperson versus an inbound lead follow-up person, give me one number to stack rank people." I worked on a pretty basic algorithm but it was 25 inputs. It spits out. You're either 99%, 100% or you're down to a 50%. It was amazing. If you look at some of the people, the top guy was Brad Munro, I remember. I'll bet you dollars to donuts. Every time I look him up on LinkedIn, he's always moving to the next level. He'll probably become a VP of sales at some point. The algorithm knows that kind of stuff. This has to be solvable.
Somebody has to do it. That's another thing. That's so exciting. There are so many opportunities, so many things we can do.
There's a time and place to do it and it seems like that time is now. Let's go into agriculture a little bit. If I'm looking at your website, it says, "Maximize the value of your post-harvest grain." You gave a little bit of an example. This batch goes bad after two months. This batch goes bad after five. Tell us the size and scale of this business problem. Are we talking about, “Here are the ten towers of grain?” Are these big, massive problems that you're solving?
They're big, massive problems. It's a worldwide problem. Problems are about $12 to $15 billion a year. This is just in the developed world. If you add China and India, it's $71 billion. The numbers are big. In the US alone, we have in corn and soybean $25 billion bushels every year in the eleven states in the Midwest. This translates to $118,000 storage locations and bins or temporary storage piles. The problem is huge. In the modern world, in the US, about 1% to 2% is lost to spoilage but in other parts of the world, it could be 14% in Brazil, 22% in Africa, 30% in India. The numbers are big. Spoil is only one part of the equation.
There are other issues like energy used. You're drying corn comes off the field with 23%, 24% moisture. If you store it at that moisture, it will be bad within a week. You drive down to 17%, 18%. If you dry it down to 12%, it'd be good for a year but what used to be 1,000 pounds now is 950 pounds. You just lost money. You don't want to over-dry it. You have a dilemma. How much do I dry? I'm fighting between the weight, which has some money or moisture, which is possible potential for spoilage? It's a tricky situation.
In some of these other places like Brazil, they're probably fudging it a little more and saying, "This one's only two months old. It's fine to sell it to the marketplace."
It gets worse because what happens is biological activity creates a hard spot. You have 500,000 bushel piles and there are some pests or moisture. Biologically, it gets bigger and bigger. If we don't cure it, don't cool, it catches fire. We see bins going up and smoke all the time. This happened on barges when they're transporting grain. Advanced knowledge through sensors and AI software when things will go bad is vital to running a clean operation, protecting the world's food supply, knowing when to turn the fans on and off to save energy. There's so much to harvest here if you're going to use the time.
AI In Agriculture: Artificial intelligence is going to call for change. Let's not fight it but use it to our advantage instead.
If AI went away or never came into existence, would you have started the business years ago? What would be different about it now without AI?
A short answer is yes. It still made sense. Think of these two buckets. Bucket number one is, "Do I have a problem?" Wireless sensors help do that, so that alone is a good business. What AI does is not just, “Do I have a problem but why did I have a problem? When will I have a problem? How do I solve the problem?” Those three questions are a more advanced benefit. If there was no AI, those would be hard to do.
You would have people running them in spreadsheets in Pakistan, for example. It would take a lot of time, energy and money to deliver that kind of data.
People like you didn't do it but people did because we have had agriculture for years. What it did were the knowledge and experience. The grandfather will open a bin, snip it and say, "It’s a little moldy. Let's do something. It sounds pretty good." That's what people did. That works half the time, if not more often. The problem is we have a labor shortage. The grandpas are retiring. The new generation is not that. They're going to need something to rescue. That's where the AI comes in.
We talked a little bit about the enterprise sales side like giving you a better forecast and those kinds of things. If you apply your thinking around what you're doing for the agriculture industry, where do you see sales headed in terms of AI deployments?
It will be a great rescue to a typical salesperson in the field. The typical salesperson is not super-efficient. They spend a lot of time prospecting, entering a bunch of data into a spreadsheet and pipeline, attending lots of meetings and then driving to meet customers. They're not very efficient. Where I see AI helping them is make them more productive by taking over some of the menial tasks and automating them. I don't know what form it will take but imagine being able to speak to an app about your last customer, what happened and whatnot. It automatically fills out a trip report, expense report and your pipeline. There'll be two steps too far but all the pieces are in place to make this happen.
That’s the middle of the funnel. I think of top of the funnel, so I did this funny Amazon Alexa thing. I don't know if I could do it. I'd have to Google it again. You can program Alexa to answer whatever question you want. It's built into the app. I did this funny video and I said, "Amazon, please pull me a list of 1,000 prospects that are 1,000 employees and higher in the medical industry." "I've done it." "Can you send an email that we know gets a 10% response rate and books at least twenty meetings with that batch of 1,000 people? Do a similar in social." You can talk to the box and then the box executes everything.
This is so beautiful. It's like a Nirvana. I don't know if you have money from that but it's happening.
It's going to be fun to see. A lot of people ask the elephant in the room question. What about all the people? I'll share some data at an AI show that I went to that showed productivity and the world's going like this. The headcount required to deliver it is going slower. What that makes me think is there had been people who've said, "I want to solve world hunger traditionally." "Good luck with that." There's a point when the productivity per head runs away so fast that you say, "Let's go solve that problem."
You're on the same wavelength as me. People ask me all the time about AI and robotics. Will there be greater unemployment and all that stuff? I am with you on that. That world I see me, I'm an optimistic one, which is a few years from now, we will be working maybe 25 hours a week, not 55, and still be able to make the same money because of the productivity boost and be able to use that extra time to do things which we want to pursue like you said, world hunger but whatever is pleasing to your community.
Entrepreneurship is a lifestyle choice. It gives you the freedom to choose what rules you want or what days you want to work.
Also, cancer or whatever it is you want to do. Think of things like universal basic income. It's pretty controversial. I generally lean fairly conservative and yet my AI thinking side is, "Let's be human. Let's solve something."
The thing is, when you give people this $1,000, they're spending it. As long as people are spending it, jobs are being created. When a job is being created, taxable income is being created. The bad thing that will happen is people put money under the mattress. If you give a millionaire an extra $100, "What am I going to do?" There's nothing different.
Americans are good at spending money, that's for sure. One hundred and twenty percent of everything they earn on average. This is a fabulous conversation. Is there anything else that you'd like to leave with the audience of the utopia or where things are headed?
AI is going to be deceptive. It's going to call change. People will be reluctant to embrace it and fight it. My point is I'm with you on this one. Let's not fight it. Let's use it to our advantage. When automobiles, electricity and computers came in, we all went through this transition. This could simplify our life. The last part, look at medical diagnostics. If you're in New York City, San Francisco, you have access to the best medicine. Imagine if you're in some faraway small town like Missouri or Mississippi. The best doctor may not be in the local hospital but with medical AI, you have access to the best, no matter where you are. That's begun to trickle down. I'm optimistic about that.
My dad was a radiologist and my brother-in-law is. They're seeing that play out. An AI can read the X-ray. My dad used to sit there. He has all the ums and uhs. "Um, clavicle number 3.4." Now, the AI can do that work in seconds, not hours and do it at a much higher level of accuracy.
There's a better world out there.
Naeem, I appreciate you taking the time and sharing your thoughts on where things are headed in agriculture, enterprise sales, the news media, all of the above. It's fabulous talking with you. Thanks for joining the show.
It's good to talk to you.
Important Links:
About Naeem Zafar
Naeem Zafar is a world citizen and the founding CEO of TeleSense, an IoT (Internet of Things) company in Silicon Valley. His last company in the enterprise mobility and security space, Bitzer Mobile, was acquired by Oracle.His career was shaped in the high-tech industry having started or worked at seven companies and advised several dozen startups. He has distilled his knowledge and personal experiences in his writings and teaching to help guide entrepreneurs all over the world. He travels frequently to lecture on a variety of topics related to starting businesses and the life of an entrepreneur.
Please visit https://westernciv.com/staff/naeem-zafar to see his lectures on history, geopolitics, and future trends.He is a faculty member at the University of California, Berkeley where he teaches entrepreneurship and strategy in the SCET [Sutardja Center of Entrepreneurship & Technology] and at the Fung Institute of Engineering Leadership.He also leads the Northeastern University’s Semester-in-San Francisco program as an adjunct professor and program director.He believes that entrepreneurship can be a powerful tool to alleviate poverty and extremism from the world and “Social Businesses” can fill the gap where public institutions often fall short.
