In a fast-growing environment, how does our small data science team continuously solve our company’s and customers’ greatest challenges?
At Razorpay, our mission is to be a one-stop fintech solution for all business needs. We power online payments and provide other financial solutions for millions of businesses across India and Southeast Asia.
Since I joined in 2021, we have acquired six companies and expanded our product offerings.
Though we’re growing quickly, Razorpay competes against much larger organizations with significantly more resources to build data science teams from scratch. We needed an approach that harnessed the expertise of our 1,000+ engineers to create the models they need to make faster, better decisions. Our AI vision was fundamentally grounded in empowering our entire organization with AI.
Given our goal of putting AI into the hands of engineers, ease-of-use was at the top of our wish list when evaluating AI solutions. They needed the ability to ramp up quickly and explore without a lot of tedious hand-holding.
No matter someone’s background, we want them to be able to quickly get answers out of the box.
AI experimentation like this used to take an entire week. Now we’ve cut that time by 90%, meaning we’re getting results in just a few hours. If somebody wants to jump in and get an AI idea moving, it’s possible. Imagine those time savings multiplied across our entire engineering team – that’s a huge boost to our productivity.
That speed allowed us to solve one of our toughest business challenges for customers: fraudulent orders. In data science, timelines are usually measured in weeks and months, but we achieved it in 12 hours. The next day we went live and blocked all malicious orders without affecting a single real order. It’s pretty magical when your ideas become reality that fast and have a positive impact on your customers.
When team members load data into DataRobot, we encourage them to explore the data to the fullest – rather than rushing to train models. Thanks to the time savings we see with DataRobot, they can take a step back to understand the data relative to what they’re building.
That layer helps people learn how to operate the DataRobot Platform and uncover meaningful insights.
At the same time, there’s less worry about whether something is coded correctly. When the experts can execute on their ideas, they have confidence in what they’ve created on the platform.
For cloud computing, we’re a pure Amazon Web Services shop. By acquiring DataRobot via the AWS marketplace, we were able to start working with the platform within a day or two. If this had taken a week, as it often does with new services, we would have experienced a service outage.
The integration between the DataRobot AI Platform and that broader technology ecosystem ensures we have the infrastructure to tackle our predictive and generative AI initiatives effectively.
In the highly regulated fintech industry, we have to abide by quite a few compliance, security, and auditing requirements.
DataRobot fits our demands with transparency, bias mitigation, and fairness behind all our modeling. That helps ensure we’re accountable in everything we do.
For smoother adoption, creating standard operating procedures has been critical. As I experimented with DataRobot, I documented the steps to help my team and others with onboarding.
What’s next for us? Data science has changed dramatically in the past few years. We’re making decisions better and quicker as AI moves closer to how humans behave.
What excites me most about AI is it’s now fundamentally an extension of what we’re trying to achieve – like a co-pilot.
Our competitors are probably 10 times bigger than us in terms of team size. With the time we save with DataRobot, we now have the opportunity to get ahead. The platform is an extreme developer productivity multiplier that allows our existing experts to prepare for the next generation of engineering and quickly deliver value to our customers.
The post Customer Spotlight: Building a Competitive & Collaborative AI Practice in FinTech appeared first on DataRobot.
The introduction of the automobile changed American culture profoundly, and it has been doing so for over a century. Innovations in automotive technology have allowed people to travel farther, faster, and for less fuel than generations before. In fact, the […]
The post 5 Ways AI Is Revolutionizing the Automotive Industry appeared first on TechSpective.
Looks like college take-home tests are destined to suffer the same fate as the Dodo bird.
Instructors at a U.K. university learned as much after a slew of take-home exams featuring answers generated by ChatGPT passed with flying colors — all while evading virtually any suspicions of cheating.
Observes writer Richard Adams: “Researchers at the University of Reading fooled their own professors by secretly submitting AI-generated exam answers that went undetected and got better grades than real students.
“The university’s markers – who were not told about the project – flagged only one of the 33 entries.”
Observes Karen Yeung, a professor at the University of Birmingham: “The publication of this real-world quality assurance test demonstrates very clearly that the generative AI tools — freely and openly available — enable students to cheat take-home examinations without difficulty.”
In other news and analysis on AI writing:
*In-Depth Guide: Lovo AI Text-to-Voice: Writers looking for a reliable text-to-voice solution may want to give Lovo AI a whirl, according to Sharqa Hameed.
Hameed’s guide on the product is extremely valuable in that it offers scores of step-by-step screenshots that truly give you a detailed look at how Lovo AI works.
Hameed’s verdict on the app: “Overall, I’d rate it 4 out of 5.
“It offers various valuable features, including Genny, Auto Subtitle Generator, Text to Speech, Online Video Editor, AI Art Generator, AI Writer and more.
“However, its free version limits you to convert up to 20 minutes of text to audio.”
*Free-for-All: Open Source Promises Wide Array of AI Writing Tools: Facebook founder Mark Zuckerberg predicts that writers and others will continue to have a number of AI choices as the tech grows ever–more sophisticated.
A key player in AI writing/chat tech, Zuckerberg has released his AI code as open-source — available to any and all to use and alter.
Observes Zuckerberg: “I don’t think that AI technology is a thing that should be kind of hoarded and — that one company gets to use it to build whatever central, single product that they’re building.”
*Grand Claims, Meh Results: Google’s AI Falls Short: Apparently, Google’s Gemini — the AI that powers its direct competition to ChatGPT — is not all it’s cracked up to be.
Observes writer Kyle Wiggers: “In press briefings and demos, Google has repeatedly claimed that the models can accomplish previously impossible tasks thanks to their ‘long context.'”
Those tasks include summarizing multiple hundred-page documents or searching across scenes in film footage.
“But new research suggests that the models aren’t, in fact, very good at those things,” Wiggers adds.
*Freelance Writing Dreams Disappearing in a Puff of Code: Add freelancers to the growing list of workers discovering that AI is less a ‘helpful buddy’ and more a ruthless job stealer.
Case in point: Since the advent of ChatGPT, job opportunities in freelance writing have declined 21%, according to a newly updated study.
Observes writer Laura Bratton: “Research shows that easily-automated writing and (computer) coding jobs are being replaced by AI.”
*Privacy Ninja: New AI Email Promises to Guard Your Secrets: Proton, an email provider long-prized for its heavy emphasis on privacy, has added AI to its mix.
Specifically, its newly released AI writing assistant ‘Proton Scribe’ is designed to help users auto-write and proofread their emails.
Observes writer Paul Sawers: “Proton Scribe can be deployed entirely at the local device level — meaning user data doesn’t leave the device.
“Moreover, Proton promises that its AI assistant won’t learn from user data — a particularly important feature for enterprise use cases, where privacy is paramount.”
*Forget Solitaire: Claude Turns AI Writing into a Collaborative Party Game: ChatGPT competitor Claude has a new feature that enables users to publish, share and remix the AI writing and other content that they generate with one another.
Observes writer Eric Hal Schwartz: “Essentially, you can open published ‘Artifacts’ created by others and modify or build upon them through conversations with Claude.
“Anthropic is pitching it as a way to foster a collaborative environment.”
*Robo Lawyer: For Many Attorneys, AI Still a Boogeyman: Despite its considerable benefits to the legal community, AI is viewed warily by many lawyers and pros.
Specifically, 77% of professionals recently surveyed by Thomson Reuters saw AI as a threat to lawyers.
Observes Artificial Lawyer: “While some very innovative lawyers are comfortable with AI and have few worries about the legal world’s imminent demise, there are plenty of lawyers out there who still feel very uncertain about what this all means for them, the profession, and their firms.”
*ChatGPT Mind-Meld: New Hope For the Paralyzed: A man slowly succumbing to paralysis has been given new hope with ChatGPT, which is enabling him to text using his brain waves.
Using a brain implant, the man is able to translate his thoughts into text commands — generated by ChatGPT — which he uses to operate computerized communication devices.
Observes the patient: “You get choices of how you might respond in several different ways.
“So rather than me typing single words, I’m hitting one or two buttons or clicks, if you will, and I’ve got the majority of a sentence done.”
*AI Big Picture: AI Gold Rush Still Runs Hot: Nearly 20 months after ChatGPT introduced a stunned world to the potential of AI, businesses across the world are still clamoring to bring the newly commercialized tech onboard.
Observes writer Ben Dickson: “Most organizations are spending hefty amounts to either explore generative AI use cases or have already implemented them in production,” according to a new survey of 200 IT leaders.
“Nearly three-fourths (73%) of respondents plan to spend more than $500,000 on generative AI in the next 12 months, with almost half (46%) allocating more than $1 million,” Dickson adds.
Share a Link: Please consider sharing a link to https://RobotWritersAI.com from your blog, social media post, publication or emails. More links leading to RobotWritersAI.com helps everyone interested in AI-generated writing.
–Joe Dysart is editor of RobotWritersAI.com and a tech journalist with 20+ years experience. His work has appeared in 150+ publications, including The New York Times and the Financial Times of London.
The post Won’t Get Fooled Again appeared first on Robot Writers AI.
A break in play during a Small Size League match.
Today, 21 July, saw the competitions draw to a close in a thrilling finale. In the third and final of our round-up articles, we provide a flavour of the action from this last day. If you missed them, you can find our first two digests here: 19 July | 20 July.
My first port of call this morning was the Standard Platform League, where Dr Timothy Wiley and Tom Ellis from Team RedbackBots, RMIT University, Melbourne, Australia, demonstrated an exciting advancement that is unique to their team. They have developed an augmented reality (AR) system with the aim of enhancing the understanding and explainability of the on-field action.
The RedbackBots travelling team for 2024 (L-to-R: Murray Owens, Sam Griffiths, Tom Ellis, Dr Timothy Wiley, Mark Field, Jasper Avice Demay). Photo credit: Dr Timothy Wiley.
Timothy, the academic leader of the team explained: “What our students proposed at the end of last year’s competition, to make a contribution to the league, was to develop an augmented reality (AR) visualization of what the league calls the team communication monitor. This is a piece of software that gets displayed on the TV screens to the audience and the referee, and it shows you where the robots think they are, information about the game, and where the ball is. We set out to make an AR system of this because we think it’s so much better to view it overlaid on the field. What the AR lets us do is project all of this information live on the field as the robots are moving.”
The team has been demonstrating the system to the league at the event, with very positive feedback. In fact, one of the teams found an error in their software during a game whilst trying out the AR system. Tom said that they’ve received a lot of ideas and suggestions from the other teams for further developments. This is one of the first (if not, the first) AR system to be trialled across the competition, and first time it has been used in the Standard Platform League. I was lucky enough to get a demo from Tom and it definitely added a new level to the viewing experience. It will be very interesting to see how the system evolves.
Mark Field setting up the MetaQuest3 to use the augmented reality system. Photo credit: Dr Timothy Wiley.
From the main soccer area I headed to the RoboCupJunior zone, where Rui Baptista, an Executive Committee member, gave me a tour of the arenas and introduced me to some of the teams that have been using machine learning models to assist their robots. RoboCupJunior is a competition for school children, and is split into three leagues: Soccer, Rescue and OnStage.
I first caught up with four teams from the Rescue league. Robots identify “victims” within re-created disaster scenarios, varying in complexity from line-following on a flat surface to negotiating paths through obstacles on uneven terrain. There are three different strands to the league: 1) Rescue Line, where robots follow a black line which leads them to a victim, 2) Rescue Maze, where robots need to investigate a maze and identify victims, 3) Rescue Simulation, which is a simulated version of the maze competition.
Team Skollska Knijgia, taking part in the Rescue Line, used a YOLO v8 neural network to detect victims in the evacuation zone. They trained the network themselves with about 5000 images. Also competing in the Rescue Line event were Team Overengeniering2. They also used YOLO v8 neural networks, in this case for two elements of their system. They used the first model to detect victims in the evacuation zone and to detect the walls. Their second model is utilized during line following, and allows the robot to detect when the black line (used for the majority of the task) changes to a silver line, which indicates the entrance of the evacuation zone.
Left: Team Skollska Knijgia. Right: Team Overengeniering2.
Team Tanorobo! were taking part in the maze competition. They also used a machine learning model for victim detection, training on 3000 photos for each type of victim (these are denoted by different letters in the maze). They also took photos of walls and obstacles, to avoid mis-classification. Team New Aje were taking part in the simulation contest. They used a graphical user interface to train their machine learning model, and to debug their navigation algorithms. They have three different algorithms for navigation, with varying computational cost, which they can switch between depending on the place (and complexity) in the maze in which they are located.
Left: Team Tanorobo! Right: Team New Aje.
I met two of the teams who had recently presented in the OnStage event. Team Medic’s performance was based on a medical scenario, with the team including two machine learning elements. The first being voice recognition, for communication with the “patient” robots, and the second being image recognition to classify x-rays. Team Jam Session’s robot reads in American sign language symbols and uses them to play a piano. They used the MediaPipe detection algorithm to find different points on the hand, and random forest classifiers to determine which symbol was being displayed.
Left: Team Medic Bot Right: Team Jam Session.
Next stop was the humanoid league where the final match was in progress. The arena was packed to the rafters with crowds eager to see the action.
Standing room only to see the Adult Size Humanoids.
The finals continued with the Middle Size League, with the home team Tech United Eindhoven beating BigHeroX by a convincing 6-1 scoreline. You can watch the livestream of the final day’s action here.
The grand finale featured the winners of the Middle Size League (Tech United Eindhoven) against five RoboCup trustees. The humans ran out 5-2 winners, their superior passing and movement too much for Tech United.