We're sunsetting PodQuest on 2025-07-28. Thank you for your support!
Export Podcast Subscriptions
cover of episode Rajarshi Gupta: Artificial Intelligence and Crypto at Coinbase

Rajarshi Gupta: Artificial Intelligence and Crypto at Coinbase

2025/2/6
logo of podcast Generative Now | AI Builders on Creating the Future

Generative Now | AI Builders on Creating the Future

AI Deep Dive AI Chapters Transcript
People
R
Rajarshi Gupta
Topics
Rajarshi Gupta: 我在Coinbase领导AI团队,负责保护每笔交易和登录安全,并通过个性化内容和通知提升用户体验。我们还利用生成式AI技术开发了员工助手和客户支持聊天机器人等应用,并积极采用Copilot、Sourcegraph Kodi和Cursor等工具来提高工程师效率。在模型部署方面,我们使用多云、多LLM平台,并通过多种方法评估模型性能,但LLM的置信区间评估仍然是一个挑战。GPU的可用性是部署AI应用面临的最大挑战之一。未来一年,我们将专注于应对潜在的交易量激增和利用AI优化公司内部运营,特别是在合规和效率方面。 此外,我认为在LLM评估、解决企业级问题以及利用生成式AI优化企业运营方面存在巨大的创业机会。区块链技术为代理AI提供了一种独特的优势,因为它能够实现可验证、不可变的交易,并支持微支付。 我们构建了多层次的防护措施,以控制模型输出、访问权限和用户行为,并保护人工客服人员。目前缺乏对LLM进行严格评估的方法,这与以往基于精确率和召回率的机器学习模型评估方法不同。 Anand Iyer: 作为Lightspeed的投资合伙人,我对Coinbase如何利用机器学习技术印象深刻,特别是其在生成式AI领域的投入和创新。 Michael Magnano: 作为Lightspeed的合伙人,我对Rajarshi Gupta在Coinbase的AI工作表示赞赏,并认为这次对话对于了解Coinbase如何利用机器学习技术非常有价值。 Eugene Chung: 作为Coinbase的长期用户,我对Coinbase在应对市场趋势方面的能力印象深刻,并对代理AI在加密货币领域的应用前景充满期待。 Josh: 作为AI职业培训平台AUX的创始人,我对LLM的置信区间评估问题以及如何将以往机器学习模型评估的严格性应用于LLM表示关注。

Deep Dive

Shownotes Transcript

This week on the podcast, we’re looking at how machine learning is optimizing how crypto giant Coinbase operates. Coinbase Head of Machine Learning Rajarshi Gupta joins the Generative San Francisco stage to discuss with Anand Iyer, Lightspeed Venture Partner focused on crypto.

(00:00) Introduction(00:26) Rajarshi Gupta’s Career Journey(01:54) Exploring Early Android Malware Detection(04:44) Machine Learning at Coinbase(09:03) Generative AI Initiatives at Coinbase(13:52) Evaluating and Managing AI Deployments(21:10) GPU Availability Challenges(23:27) Workflow with Specific Hosted Instances(25:43) Coinbase: Scaling and Future Plans(29:09) Opportunities for AI Startups(31:46) Audience Q&A: Data Gap with AI Startups(34:48) Audience Q&A: Agentic AI and Crypto Integration(38:26) Audience Q&A: Guardrails and Frameworks for AI(42:43) Audience Q&A: Evaluating LLMs(45:40) Closing Thoughts

Stay in touch:www.lsvp.com)X: https://twitter.com/lightspeedvp)LinkedIn: https://www.linkedin.com/company/lightspeed-venture-partners/)Instagram: https://www.instagram.com/lightspeedventurepartners/)Subscribe on your favorite podcast app: generativenow.co)Email: [email protected])The content here does not constitute tax, legal, business or investment advice or an offer to provide such advice, should not be construed as advocating the purchase or sale of any security or investment or a recommendation of any company, and is not an offer, or solicitation of an offer, for the purchase or sale of any security or investment product. For more details please see lsvp.com/legal).