At tasq.ai, we believe that excelling in AI requires addressing two major challenges: data and power consumption. #Human #guidance is essential in overcoming the data challenge, ensuring our AI solutions are both effective and efficient. Recent incidents, like Microsoft's AI mistakenly accusing a reporter, underscore the need for robust AI oversight : https://lnkd.in/dg-S-2KH Even the most successful models require a human layer to ensure responsible and ethical AI. Interested in learning how tasq.ai can help you navigate the complexities of AI with #human-#guided insights? Contact us today
tasq.ai
Software Development
Tel Aviv, Israel 3,769 followers
The Science of Human Guidance for AI
About us
Tasq.ai is the first NanoTasking Data Platform providing data creation and curation by dynamic mass allocation of global, diverse human judgements. We developed the only solution that can provide Human Guidance for AI at mass scale without depreciation of quality. Tasq.ai's solutions are built ground up to offer Fairness, Diversity, Equality and dedicated to support Responsible AI. We empower Data Science and ML teams to create high quality training datasets for AI developments by breaking down complex sceneries to numerous nano-tasks for human annotation and verification. Our unique method allows our customers to enjoy an upgraded experience in scale and quality to any use case from any vertical.
- Website
-
https://www.tasq.ai
External link for tasq.ai
- Industry
- Software Development
- Company size
- 11-50 employees
- Headquarters
- Tel Aviv, Israel
- Type
- Privately Held
- Founded
- 2019
Locations
-
Primary
Tel Aviv, Israel 6777016, IL
-
New York, US
Employees at tasq.ai
-
Nati Catalan
Co Founder & CRO at Tasq.ai | Generative AI, Machine Learning, Data Science
-
Yossi Motro
Co-Founder & CTO at tasq.ai
-
Konstantin Kraimer
Senior Technical Lead at Tasq.AI | Ex-Amazon
-
Asaf Katz
Senior Frontend Developer & UX Designer @ tasq.ai | Remote Worker | Open Source Developer
Updates
-
We're eager to share the latest research from our Chairman, Prof. Shai Dekel: "Wavelets Are All You Need for Autoregressive Image Generation." This innovative paper presents a fresh approach to image generation, utilizing #wavelet coding to create visually striking, high-quality images with a focused and effective method. Explore how this could shape the next era of #AI and image #generation. Dive into the full research here : https://lnkd.in/dJw7gzth At tasq.ai, we're continuously working on improving #Gen #AI #evaluation solutions. As AI technology advances, we remain committed to providing accurate and reliable evaluation processes. Subscribe to our newsletter to stay updated! https://lnkd.in/dYEBWb-G
-
-
tasq.ai has been handpicked to join the #MindTheTech London 2024 Dream Team by Calcalist כלכליסט! In times like these, being chosen truly energizes us. We’re proud to stand among Israel’s leading startups and share our innovations with the world. Mark September 16-17 on your calendar and join us in London! With valuable networking and opportunities to connect. Secure your ticket now on the conference website: https://lnkd.in/gUFNaQKF
Thrilled to announce that tasq.ai has been selected by Calcalist כלכליסט as a member of the prestigious #MindtheTech London 2024 Dream Team! In these challenging times, being part of this conference is a testament to tasq.ai's resilience and potential. We're excited to join Israel's most promising startups at this global tech showcase. Can't wait to connect with top-tier investors and industry leaders as we make our mark on the global stage! Join us in London on September 16-17 for two days of inspiring talks, game-changing ideas, and endless networking opportunities. Secure your ticket now on the conference website: https://lnkd.in/gUFNaQKF #MindTheTech #DreamTeam #tasqai #London #techconference #investment #innovation #IsraeliTech The Dream Team: Amiram Shachar | Michal Tzuchman Katz, MD | Gilad Yarkoni | Tzvika Perelmuter | Daniel Adler | Alex Gold | Adiel Am-Shalom | Boaz Nol | Cohen Daniel | Omer Goldschmidt | Nati Catalan | Yuval Klein | Ron Ganot | Dr. Lihi Raichelson | Anat Natan | Benjamin (Benny) Avraham | Adi Baltter
-
-
Boosting Ads Integrity with tasq.ai Maintaining the integrity of #ads is crucial in digital advertising. Misleading content, inappropriate visuals, and false claims can harm brand reputation and user trust. That's where tasq.ai steps in. Here’s how we help: 1. Content Check: Our platform reviews ad content to ensure it meets industry standards and regulations, catching misleading claims and inappropriate content. 2. Visual Review: Using computer vision, tasq.ai flags visuals that don't meet brand guidelines or legal requirements. 3. Text Analysis: Our language models scan ad copy to correct any misleading or false statements. 4. Global Human Feedback: We use human feedback from diverse locations worldwide to validate ad content, ensuring cultural sensitivity and compliance with local regulations. 5. Scalable Solutions: Our AI and human feedback approach handles large volumes of ad content efficiently, scaling with your needs. By leveraging #tasq.ai, companies can ensure their ads are effective, ethical, and compliant. Check out this recent article on Google's efforts to block bad ads and improve ad safety with AI: #Google Blocks 5.5 Billion Bad Ads, Leverages AI for Improved Ad Safety. (link in the first comment) Let’s work together to create a safer, more trustworthy advertising ecosystem. #AdsIntegrity #AI #Compliance #DigitalAdvertising #BrandSafety #tasqai #AIInnovation #HumanFeedback #Scalability
-
-
We are excited to join forces with Accenture on GenAI and technology innovation. Don’t miss out on Ehud Barnea, PhD talk on optimizing GenAI with human-as-a-model! See you there! Please note – the number of places for the physical meeting is limited. #genai #innovation #AI #data
בעולם אידיאלי הכל קל ופשוט אבל החיים, לצער חלק ולשמחת אחרים, הרבה יותר מורכבים. ולכן הכנו עבורכם את המפגש ההיברידי (פיסי ומקוון) הבא במסגרת מרכז מצוינות DATA & AI של The Israel Chamber of Information Technology - הלשכה לטכנולוגיות המידע בישראל : "לשחק בליגת העל של Gen-AI - איך להוציא את המקסימום ממודלי שפה". אני מזמין אתכם להצטרף למפגש מרתק עם מרצים מעולים שיעסוק באסטרטגיית פיתוח יישומים מבוססי Gen AI בפריזמה של הארכיטקטורה הטכנולוגית, הכלים והגישות הרלוונטיים. אנו נתמקד באופן שבו נכון לשלב טכנולוגיות בתחום וגישות שונות על מנת לאפשר לארגונים לפתח מענה ייעודי, מותאם , באיכות גבוהה וביעילות לצרכים מורכבים המפגש ההיברידי יתקיים ביום חמישי ה- 25.7.24 בשעה 15:00-16:30. משתתפי המפגש הפיזי מוזמנים להגיע להתכנסות שתחל בשעה 14:30 במשרדי חברת אקסנצ'ר (בניין B קומה 6) .Accenture Israel המנופים 2 הרצליה פיתוח (חניה בחניונים הסמוכים) קישור לרישום למפגש : https://lnkd.in/dXkGFzHi בטופס הרישום אתם מתבקשים לציין "פיזי" או "מקוון" נתראה! שימו לב – מספר המקומות למפגש הפיזי מוגבל. ההשתתפות מותנית בקבלת מייל אישור בסמוך למועד המפגש. עדיפות תינתן לחברי לשכה רשומים. Dan Klein, Ehud Barnea, PhD Michael(Mike) Erlihson Nir Makover Avishay Meron Boaz Shafrir Yossi Levi 🇮🇱 Claudia Levy Noa Franko-Ohana Accenture Israel, tasq.ai, Stealthy #GenAI #LLMs #KnowledgeGraphs #Finetuning #Automation #HumanAsaModel
-
-
tasq.ai is now live on the Microsoft Azure Marketplace! Our Ultra-Scale Human Guidance Services for #ML #evaluation and training are designed to help businesses achieve unparalleled accuracy and efficiency in their #AI and machine learning projects. By leveraging our groundbreaking technologies, companies on Microsoft Azure Marketplace will be able to set new standards for processing data for AI. What does this mean for you? Seamless Integration: Easily integrate our service with your existing infrastructure. Scalability: Handle large volumes of data effortlessly. Accuracy: Achieve superior data quality with our innovative approach. Efficiency: Speed up your AI projects and go to market faster. Join the growing number of businesses transforming their AI data processes with tasq.ai. Check out our service on Microsoft Azure Marketplace https://lnkd.in/dDV3iQ8R
-
-
CVPR last week exceeded all expectations! We can't wait to see where the future of AI takes us and to continue these impactful discussions and collaborations. #CVPR2024 #AI #ResponsibleAI #Tasqai #Innovation
VP Partnership & Business Development | Innovation | Team builder | Product & Technology | Strategy & Partnerships | Cloud and GenAI
CVPR last week was mind-blowing! Tasq.ai truly made an impact, and I’m excited to share some of the key moments: 1. Workshop Success! Our Head of AI, Ehud Barnea, PhD, led a fascinating workshop on retail innovation with Amazon. The session, titled "RetailVision - Field Overview and Amazon Deep Dive," was insightful and forward-thinking. It was a must-attend for anyone interested in the future of retail! Ehud's Workshop (https://lnkd.in/dMhcwaKv) 2. Panel on Responsible AI - I had the honor of being part of a distinguished panel discussing Responsible AI. We dived into critical conversations about the ethical implications and responsibilities that come with advancing AI technologies. My Panel on Responsible AI (https://lnkd.in/dSXQYTEH) 3. Partnerships- We had meaningful conversations with Centific, Voxel51, Tenyks, Meta, Sony, and more! 4. Networking Opportunities - Side events, happy hours, workshops, and keynotes provided many opportunities to connect around emerging trends and discuss how Tasq.ai can elevate AI workflows. It was fantastic to connect with so many industry leaders. Looking forward to continuing these important conversations and exploring the future of AI together! Botty Dimanov, PhD Caner Hazirbas Susanna Ricco Emily McReynolds Shiran Gan-Or Yosi Keller Riyana Patel Harpreet Sahota 🥑 Prerna Dhareshwar Shiva Jayaraman #CVPR2024 #AI #ResponsibleAI #RetailInnovation #Tasqai #Networking
-
-
-
-
-
+3
-
-
Today, we are excited to share that tasq.ai and Tenyks have partnered to offer a comprehensive solution for management, curation, training and evaluation of #AI and #computervision based applications. Tenyks's visual data management platform powered by Intelligent Search together with tasq.ai's global diverse human guidance target to provide enterprise customers with a complete state of the art solution to improve accuracy while cutting time to production and total cost of ownership. tasq.ai's on-demand global networks consisting of millions of qualified human taskers, coupled with Tenyks ability to chain and fine-tune state-of-the-art foundation models, creates the ultimate method for extensive #data exploration, curation, model #evaluation and fine tuning. Together, our customers will profit from extracting actionable insights from unstructured visual information at massive scale, faster and more cost effective than ever. Want to find out more? Visit our website now : www.tasq.ai Noa Franko-Ohana Botty Dimanov, PhD Nati Catalan
-
-
Our VP of Partnership, Noa Franko-Ohana will be participating in an upcoming panel on #responsible #data, which is part of the #CVPR conference, showcasing our commitment to ethical AI practices. #AI is revolutionizing our world, but with great power comes great responsibility. At Tasq.ai, we're not just keeping up; we're leading the charge in responsible AI, ensuring our solutions are ethical, inclusive, and impactful. What sets us apart? Our diverse, global community. Spanning continents and cultures, we include individuals of all ages and backgrounds. This diversity is our foundation, allowing us to develop AI with a deep, nuanced understanding of the world's complexities. Imagine #AI trained not just by data scientists in tech hubs, but by people from rural villages, bustling cities, and everywhere in between. Each perspective adds richness to our AI, making it more robust and fair. Our technology’s global reach ensures that AI systems are built with empathy and insight, tapping into a wide array of perspectives to mitigate biases and create equitable solutions. Consider the challenge of #ML model evaluation. When models are evaluated by a homogeneous group, biases can emerge. Our diverse #crowd ensures that our AI models are trained on data reflecting a multitude of experiences and viewpoints, significantly reducing this risk. Moreover, we integrate #human judgment into our AI development process. This human-in-the-loop approach enhances accuracy and reliability, allowing us to catch and correct potential biases early on. At Tasq.ai, our commitment to responsible AI is unwavering. We prioritize transparency, fairness, and privacy, ensuring that our technology benefits everyone. Join us in building a future where AI serves all of humanity, reflecting the true diversity of our world. https://lnkd.in/d5uK2QZd
CVPR 2024 Workshop on Responsible Data
responsibledata.github.io
-
The Critical Need for Thorough AI Evaluation - A The Wall Street Journal test of five leading AI chatbots (OpenAI’s ChatGPT, Microsoft’s Copilot, Google’s Gemini, Anthropic’s Claude, and Perplexity) showed that no single chatbot excels in all areas. (link to their great article in first comment) This highlights the importance of comprehensive testing before making business decisions. The WSJ used the same group to create and judge prompts, ensuring consistency but also introducing potential bias. A single group's evaluation reflects their specific perspectives and biases, which can skew results. Limitations of WSJ’s Evaluation Approach: 1. Homogeneous Perspective: Limited diversity of thought can lead to a narrow view of what constitutes a "good" response. 2. Potential Bias: Evaluators’ experiences and knowledge can influence their assessments, leading to subjective evaluations. 3. Lack of Real-World Variation: A single team may not represent the varied needs and questions of different users, resulting in an incomplete evaluation. Why tasq.ai Leads the Way: 1. Unmatched Expertise: Thousands of evaluators, selected for their cognitive abilities and domain expertise, provide deep and broad insights. 2. Superior Speed and Precision: Faster and more accurate evaluations help businesses make informed decisions quickly. 3. Global Perspective: A diverse evaluator base enhances the reliability and relevance of assessments. Before committing to any AI for your business, ensure it has been tested and evaluated. Tasq.ai’s advanced evaluation methodology guarantees you get the best-fit AI for your needs. #llm #AI #genAI