Vivian Chong’s Post

Story of the Week: From Hackathon Success to Personal Impact! 🚀 The first two days of this week were nothing short of exhilarating as I immersed myself in my first Trust and Safety Hackathon. Proud to be a part of Team 23 (Vivian Chong, Marcelo Davila Gonzalez, Chandana M.S. ), we put our heads together and secured 3rd place in the Trust & Safety Hackathon. Our idea on Standardizing Trust and Safety Evaluation for Large Language Models was a standout, earning us some rave reviews from Jonathan Bellack. Thank you so much for the reviews! The best part of the hackathon wasn't just the success, but the incredible ideas and collaborative spirit that filled the room. The discussions and ideas exchanged during the hackathon resonated deeply with me, reinforcing my respect and passion for the role of trust and safety in our lives. Working in the trust and safety domain has shown me the profound impact our work can have on people's lives. From protecting user information to combating identity fraud and payment fraud, every effort in this field contributes to a safer and more secure online environment for all. For anyone interested in the space, have a look at this repo:https://lnkd.in/g7eWcMwi based on Andrew Ng's course.

As Christmas approaches, online shopping activity is on the rise, and so is the risk of financial fraud. As a social enterprise, we want to draw attention BrokerChooser's ScamVestigator – an interactive game that helps you spot scams. We believe that if we can identify fraud in a game, we’re more likely to recognise it in real life. Check it out here: https://lnkd.in/dG4BEArf A big thank you to BrokerChooser for creating this valuable tool in the fight against fraud. Like, share, and let's work together to raise awareness and prevent fraud! 🤞

Join TD Bank, Fierce Urgency of Now Festival - City Awake, a program of the Greater Boston Chamber of Commerce, and Justin Cross for an informative seminar on fraud prevention called TD Safe. In today’s fast paced world, protecting your financial information is more important than ever. That’s why TD Bank is hosting this seminar to help you stay safe. Don’t miss this one! #TD Safe #proudtobeTD

**HACKED!** It was a wonderful experience working with my teammates Xinhui Li and Samin Mahdipour to win the first MEXA hackathon! Thanks to MEXA for putting on an exciting challenge and supporting us with highly engaged technical consultants and lived-experience experts. Our goal was to leverage LLMs to offer psychiatrists objective, behavioral measurements of inattention in support of their clinical ADHD diagnoses. To accomplish this goal, we tried to generate realistic transcripts of ADHD patient interviews (using single-shot learning with LLM) and then use these transcripts to quantify attentional lapses (e.g., errors of omission, commission, distraction, and perseveration) using a second LLM prompt. Here are a few takeaways from my experience: 1. Our LLM can reliably generate realistic interview content (where patients describe symptoms of inattention). 2. But our LLM struggles to simulate a skeptical psychiatrist who presses, prompts, and challenges patients to better understand the significant consequences of their inattention. 3. And the LLM struggles to simulate (and accurately quantify) realistic, inattentive response behavior in patients. #mentalhealth #AI #Hackathon #ADHD #psychiatry #googlehealth #deepmind #MEXA #Wellcome

I'm happy to share that our team won 3rd place in the MEXA hackathon on the intersection of AI and mental health.🎉 We built a solution that periodically checks in with individuals living with mental health issues by gathering their experiences from passive data, such as video, audio, and text. This data will be analysed and combined with the strengths of generative AI models like Google Gemini Flash-1.5 to provide clinicians with deeper insights into their clients' mental health between therapy sessions. This holistic measurement is more reliable compared to traditional assessment methods such as clinical interviews and questionnaires, which only captures a snapshot of someone's mental state. Our solution can: ✅Analyse text messages from check-ins, journal entries, messages etc ✅Understand emotions, vocal expressions, and speech prosody from clients' speech ✅Detect emotions based on facial expressions using Py-feat open source toolbox ✅Generate comprehensive reports about the client's mental health by synthesizing insights from video, audio, and text data. It was a pleasure working with my amazing and talented team Jamie Ferguson, 🏄♂️ Liam Swift, Harry Ross, Jack Glenn and Sandra Garrido, PhD. Thanks to the mentors from Google DeepMind, Wellcome Trust and Neuromatch

Completed the KrakenKeylogger Blue Team Lab at CyberDefenders, where an employee sought help with a task and received a finished assignment from someone who demanded payment, indicating a threat actor. The digital forensics team was then tasked to analyze several aspects, including the web messaging app used by the employee to communicate with the attacker, the password for the protected ZIP file sent by the attacker, the domain used by the attacker to download the malware's second stage, the command injected by the attacker using LOLAPPS for persistence, the complete path of the malicious file used for persistence, the application used for data exfiltration, and finally, the IP address of the attacker. CyberDefenders #cyberdefenders #karkenkeylogger #blueteam

Learning more about various businesses and technology can be facilitated by taking part in hackathons and interacting with others. Discussions with other participants, mentors, and sponsors provide you with fresh perspectives on emerging patterns, successful tactics, and original approaches to problem-solving. Networking gives you the chance to exchange ideas, ask for guidance, and form priceless connections that will serve you well long after the hackathon is over. It also helps you form relationships and locate assistance within the community.

Happy to share that our team won the 2nd place at MEXA Hackathon! 🎉 Our challenge was to use Large Language Model (LLM) & AI to overcome the challenges we face with mental health measurement. Over three intense days, we developed a working prototype that: ✅ Converts speech to text ✅ Extract emotions from the transcription. ✅ Generates a score on a clinical questionnaire used to assess patients with postpartum depression ✅ Creates a chatbot that adapts its responses based on symptom analysis and questionnaire scores Our solution aims to support patients directly and empower clinicians with AI-driven tools for better care. We’re committed to taking this project to the next level, refining our solution, and exploring its potential to make a meaningful impact :) Let's see what we can achieve with this beautiful and talented team! Viviana Greco, PhD, Gary Brown, Tolulope Oladele & Kate Sherratt Thanks to our mentors from Google DeepMind ! #AI #MentalHealth #Neurosciences #PostPartumDepression

Our experts in Osborne Clarke - Sweden have advised Detectify, a leading Attack Surface Management platform for AppSec teams powered by elite ethical hackers, on the acquisition of a majority stake in the company by global software investor Insight Partners. The transaction, finalised in October 2024, provides new funding to accelerate product innovation and drive profitable growth. The team that advised Detectify was led by corporate partner Mikael Nelson and senior associate Beatrice Suomela Read the full story on our website > https://lnkd.in/ebKTtqGj

If you missed the action last week - you can now watch the recording of “Leveraging Threat Actor & Red Team Infrastructure for Intelligence” with Jacob Williams & William Elchert on our YouTube Channel 🎉 #cybersecurity #threatintelligence #redteam #PacficRim #Sophos