Design Blog
Thoughts & Musings
I was recently really humbled to be featured at the age of 20 by Los Angeles Business Journal as one of their 20 in their 20’s business leaders in the year 2020 (thats a lot of 20’s).
Everyone on the list is doing some amazing things and I’m really lucky to be featured among this amazing group of young leaders.
Check out my interview with Mediha DiMartino of the LA Business journal below:
https://labusinessjournal.com/news/2020/jun/15/20-in-their-20s-joel-joseph-kaushal-saraf/
My friend Michael and I after the USC Demo Day Pitch Competition.
Last week I had the amazing opportunity to showcase my startup Atomus at the University of Southern California’s Demo Day event. The event brought together over 500 cofounders, investors, and students from the Los Angeles community together under one roof to show the amazing startups USC has produced.
Check out this article about it here: https://dot.la/usc-demo-day-2020-2645075471.html
One amazing thing about this event is that it really brought together the Trojan Family. Many times when working on a startup its really easy to get buried in the trenches with work. Events like Demo Day give you the feeling that you are not alone in building your startup and there’s a family of others around you willing to help.
Tim Ellis, the cofounder of Relativity Space, and I after his keynote speech.
One major highlight of the event was the key note speech by Tim Ellis, cofounder and CEO of Relativity Space. Relativity 3D prints rockets which is absolutely crazy. During his keynote he talked about his experience as a student entrepreneur starting his company including a crazy story of how he got his first investor, Mark Cuban, through a cold email. He then presented the most recent deck Relativity had used when it had closed its 140 million dollar Series C round.
One thing he emphasized during the presentation was how influential going through YCombinator, the famed startup accelerator, was in his journey building his company. After the presentation when I had the opportunity to pitch my company to him he was so impressed at the problem our team was solving in the 3D printing space that he offered to write me a letter of recommendation to YCombinator. This was probably the highlight of the day given how involved in the 3D printing space Tim is.
Some other major highlights from the event included seeing other startups in the USC ecosystem. My 3 favorites from the event were:
1) Hatch Credit - A free way for college students to improve their credit scores using existing subscriptions like Netflix, Spotify, and rent payments
2) Rally - A social media platform that allows users to be directly connected with the social issues that they care about.
3) Volv - A modern news app that consolidates each story into 9 seconds. This is so well designed I’ve been using it on my iPhone everyday since Demo Day.
Overall I had an amazing time at Demo Day 2020. Special thanks to the amazing team at Troy Labs and James Bottom for putting this event together.
I had an amazing experience getting to pitch my startup, Atomus, on stage at the 2020 Consumer Electronics Show in Las Vegas. It was amazing being one of the youngest people, if not the youngest person, on stage at the conference.
It was absolutely crazy walking through the aisles of technology companies presenting their technology. I actually got lost a couple times inside the convention halls given the sheet size of the buildings.
I had an amazing time at the USC Stevens Student Innovator Showcase this weekend. Now in its 13th year, the USC Stevens Student Innovator Showcase is held annually during Trojan Family Weekend. Student teams earn a spot in the Showcase based on their demonstrated success at creating scholarship with consequence by coming up with creative, viable solutions to global problems. During the daylong event held at USC’s University Park Campus, student teams present and pitch their innovations, inventions, and startup ideas in all disciplines for awards that help them develop their ideas further.
I had the amazing experience of talking to a host of experts from patents, technology, and commercialization proving why my team and technology was pioneering the most innovative technology at USC. More over the judges got an inside the scenes look to some of the really cool applications of our technology with the military and commercial industry, as well as an insiders look into our team’s technology and why exactly our patents and trade secrets on our core technology are so disruptive to the industry. Some of the distinguished judges who are extremely notable in the field my team is disrupting and I had the pleasure of speaking to and fielding questions from included:
Dr. Randolph Hall - Dr. Randolph Hall leads USC’s $900 million annual research enterprise, overseeing research advancement, administration, and ethics. Hall’s experience includes serving as the founder for two national research centers, the Center for Risk and Economic Analysis of Terrorism, and the National Center for Metropolitan Transportation Research. Dr. Hall is Professor in the Epstein Department of Industrial and Systems Engineering. He is the author of Queueing Methods for Services and Manufacturing. Hall served as chairman of the University Industry Demonstration Partnership, the national leader in developing standards for industry funded research in universities.
Julia S. Metcalfe - Previously Julia spent 8 years at Facebook. She spent her last year at the company spinning up a product team in the Applied Machine Learning organization, focused on democratizing data annotation to meet the needs of the ever-scaling use of machine learning across the company. She spent her first 7 years at Facebook leading the Product Experience Analytics organization for monetization products. She founded this team and scaled it to 60 people across 5 offices, evolving the discipline from operations to analysis and product management. Her team changed how Facebook understands fundamental advertiser needs by developing data-driven insights that help focus the organization on building trustworthy products.
Dr. John Wetherell - John Wetherell, Ph.D., is head of Pillsbury’s San Diego Intellectual Property group and firm co-chair of the National Life Sciences Group. Dr. Wetherell has 25+ years’ experience closing IP deals and securing U.S. and international patents for life science companies. He handles IP acquisitions, licensing and transactional due diligence; prosecutes patents; performs patent infringement and validity analyses; freedomto-operate opinions; and provides strategic IP counseling. Dr. Wetherell has written and lectured extensively on patent law, strategy, and related transactional issues for many organizations and institutions around the world.
Along with a dozen or so other judges. Fielding questions from these experts in the field really made me think on my feet and respond back with some of the exciting reasons our team was excited about our technology. Overall I had an amazing time at the event.
Me playing with one of the really cute puppies at the shelter
I’m a huge dog lover, like the biggest dog person you’ve ever met. And as a big dog person, the biggest problem for me is that I have no dogs at home. That’s why I decided to go to the Humane Society this summer and see how hard it would be to possibly adopt one.
Now one major thing that working on startups has kind of engrained in me is a lot of lean startup methodology, specifically when it comes to customer discovery. And its led me to always want to understand more about everything around me and how it works, not specifically looking for problems but more like stumbling across them. In the process of understanding how adoptions work, I had the opportunity to talk to 6 different staff and volunteers just casually asking them about their jobs and experiences. As I spent more time just talking to them, each of them individually spoke about different problems and pain points in the adoption process. At the end of the day, I not only walked away with a huge amount of respect for all the volunteers and staff, but also a fundamental problem shelters were facing: no one knew what types of dogs they were giving to families.
I learned that a lot of the dogs that come through my local shelter are from rural communities which many times don’t have a lot of background history associated with them. When the shelter takes dogs in from literally anywhere, they have a staff member basically “eyeball” what type of dog they are dealing with. As a result there is no way to actually know which type of dog people are adopting. Now one common solution to this could be genetic testing each dog that comes through, but at 100 dollars a genetic test for a dog its a non starter for a lot of non profits who don’t have the budget for it. Not to mention the month it would take to get the results back, during which these dogs would not be able to have a home and stuck in the shelter.
Me with one of the volunteers at the shelter
After going home I decided to do some additional research and find if this was an isolated problem. And it turns out, my local shelter is not alone at all. In fact at many shelters dogs are misclassified into wrong breeds, and the result is they take longer to get adopted or face a higher rate of returns back to the shelter.
Now one of the really amazing things about being a Cognitive Applications Technical Intern is I get to work with some of the most advanced and cutting edge technologies that run on IBM’s cloud. And when you work 40 hours a week on something that you think is really amazing and cool, your problem solving mindset does just magically switch off after your normal 9-5 job. So I came up with an interesting idea: What if instead of a human “eyeballing” what breed a dog was we had IBM Watson do it instead.
Through a combination of open source data sets I was able to build a custom machine learning model that was able to determine what guess what breed a dog was. Now one thing that really goes understated a lot of time is the amount of time you need to find getting, cleaning and optimizing the right data. I some really great learning experiences trying to do a lot of the data wrangling there and even learning how to write some scripts to automate the process.
After I built the model, I made a simple iOS app for volunteers and staff at the shelter to use it. The UI is pretty simple since this is all a MVP. But essentially you open the app, take a photo of a dog, and you get the results of what breed the model thinks it is. Right now the model I built is trained on about 20,000 dog photos and around 120 or so different breeds.
A screenshot of the MVP running on my phone.
With a MVP in hand and research and customer discovery interviews done I was able to go back to my local dog shelter and propose this project I built. I made a short presentation to their leadership of the shelter which is attached below.
The leadership of the shelter loved the idea, and currently I’m in the process of making some improvements on the app and model in order to conduct a pilot program to see the actual impact this has on finding these cute amazing doggies a home.
I had a great time working with a couple of the other SF interns for the IBM intern hackathon. One of the amazing things about this project was that our team chose to not only build a project that had a strong business case for IBM, but also had a strong social impact for the visually impaired who use the internet.
We made a video of product demo. Feel free to check it out below:
We also made a brief overview handout for the judges since they were all IBM executives. The results really speak for themselves, check it out below.
Our team also worked on making some really funny photos to post on our Team Twitter account to show just how cool working at IBM was. Here is my favorite of our team bowing down at the IBM Watson desk in our office.
Today I had the great experience of going to the Tenderloin district with a couple of my fellow interns after work and spending the evening volunteering. As a coder I pride myself on my physical ability. After-all coding for hours a day means that my hand muscles are in tip top shape. I was able to use those skills not only for coding today, but to serve almost 450 hot meals at a homeless shelter today.
One of the benefits working in the Watson West building in SF is that my building holds the Watson Experience Center. “The IBM Watson Experience Centers are permanently branded experiences, with locations in New York City, San Francisco, and Cambridge. The centers are invitation only and their primary audience is potential customers of IBM Watson.”
It was amazing going inside the center where they immerse you with the real world use cases of IBM technologies and the inside secrets the presenters use in order to wow CEOs of major companies who come in for a visit. Sadly the facility doesn’t allow photos inside so I can’t show any of the cool stuff that happened inside the room.
This week I had an amazing time exploring other IBM offices all around the bay area, not just in SF where I am based. I had a great time meeting Dinesh Nirmal and Nimesh Bhatia and spending the day at IBM Silicon Valley Lab with my fellow interns in the morning. Dinesh is the site executive of Silicon Valley Lab and is the VP of AI and Data at IBM. I was part of a group of interns who got to have a meeting with him and ask him questions about the future of IBM in the cloud and in enterprises.
After exploring SVL, we went over to the Almaden Research Lab and saw all the amazing work happening with quantum and neuromorphic computing under Jim Spohrer, Kevin Roche, and Myron Flickner.
This weekend I had the amazing opportunity of being able to help mentor hackers and represent IBM at Angel Hacks Silicon Valley. It was really amazing being able to spend the day on the floor of the hackathon and meet and help developers who needed help troubleshooting their code or just wanted to learn more and ask questions about the cool stuff that IBM was working on.
This 24 hour hackathon saw over 120+ competitors come together to form more than 40 teams at School 42, an alternative tech school in Fremont which offers completely free tuition and housing to their students. Unrelated, but this school had probably the best hackathon food I’ve ever had.
I’m really thankful IBM offers their technical interns the amazing opportunity to help out at these types of events. For me its really important to give back and help educate the next generation of hackers given that I’ve learned so much in the hackathons I’ve participated in.
One of the really nice things that I liked about this hackathon was the fact that they allow each of their sponsors to set up different objectives for what they want the projects to address. IBM in specific wanted hackers this year to work on projects related to Call for Code, which is essentially an investment not only in millions of dollars but also IBM staff and resources to help create better responses to natural disasters. Specifically in this hackathon IBM wanted developers to work on “an application using cloud, data, analytics, AI, IoT, blockchain to improve disaster preparedness, response, recovery or building back better focused on the health and well-being of individuals and communities.”
Specifically 3 projects that I really really liked were:
Advancers AI - This team from Myanmar was able to create a smartphone application that is able to identify malaria more accurately and with greater throughput than the current rapid malaria test which is currently done in 3rd world countries. They were able to get actual training data from hospitals in Myanmar and have a 97% accuracy using their smartphone. I thought this one was the coolest. Also as someone who had malaria when I was in India, I have a personal experience knowing how horrible it is.
ML Matter - This team wanted to help big disaster relief organizations in both the private and public sector access to data that could help speed up their response and help them evaluate where to send the most resources in the aftermath of a natural disaster. Their application takes images before and after a natural disaster and is able to determine which pieces of infastructure: roads, bridges, buildings, etc are down.
AIDOC - This team was able to create an AI triaging system which could have possible uses in the aftermath of a natural disaster. In the hackathon the team was able to implement two different types of tests using AI. The first was a vision test which could be used to diagnose eye problems. The second was a skin analyzer which was able to tell if certain rashes or spots on a person are a major cause for concern.
