Articles by Anshu
Activity
46K followers
-
Anshu Sharma shared thisAI training data has become a massive $12B+ business growing at an exponential growth rate. If we want AI to be able to read financial reports, or help with medical transcripts - we need to give them good training data. A lot of companies have emerged in this space and grown to massive sizes. We are working with many of the top firms in this business, and their businesses are evolving. The initial buyers were the top 2 model builders - OpenAI and Anthropic. Then, came the Neo labs - each one with billions of dollars in funding and a desire to train models with more and more data. Next, enterprises that want to train or fine tune models for their own specialized use cases. Each wave comes with demand for more data. More data comes with more risk - you don't want your training data to contain confidential or personal information - this is much harder than it looks. Because you can't just remove all such information without destroying the knowledge. The key is to preserve the shape of the data including referential integrity. We have been doing this for the leading AI companies for last 3 years. If you want to work with us, ping me or anyone at Skyflow.
-
Anshu Sharma shared thisThere's a few people who laid the path for people like me to follow. One of them is Kanwal Rekhi, the first Indian American immigrant to take a company public on Nasdaq. Just read his book - a memoir of his journey from an IIT engineer who arrived in this country, much like many of us, with a few dollars in the pocket and completely clueless about everything except one thing: we knew science and technology, and could learn almost anything. I wasn't sure if I would get anything from the book since I was vaguely aware of the story - but I was pleasantly surprised. Literally could not put this book down once I started (audible) reading it. Highly highly recommend it. Kanwal will inspire you, and make you fall in love with life (and his life story). https://lnkd.in/g3kQtqRGThe Groundbreaker: Entrepreneurship, the American Dream, and the Rise of Modern IndiaThe Groundbreaker: Entrepreneurship, the American Dream, and the Rise of Modern India
-
Anshu Sharma reposted thisToday, we are launching Skyflow for Glean: Runtime data control layer for securing the context both at ingest and inference, with in production live customers like Kearney who want to deploy AI to every employee while securing sensitive customer data at the fine-grained word/token level. "We don't want to be in the data sanitization business," said Mark Johnson, Partner and Chief Information and Digital Officer at Kearney. "Skyflow gave us per-field encryption, runtime policies, and customer-specific isolation as a platform – not a project." We protect data-in-use so you can safely run agents. If you are deploying AI in your company, we would love to talk to you. We will be at the Glean Go conference. cc Arvind Jain
-
Anshu Sharma shared thisToday, we are launching Skyflow for Glean: Runtime data control layer for securing the context both at ingest and inference, with in production live customers like Kearney who want to deploy AI to every employee while securing sensitive customer data at the fine-grained word/token level. "We don't want to be in the data sanitization business," said Mark Johnson, Partner and Chief Information and Digital Officer at Kearney. "Skyflow gave us per-field encryption, runtime policies, and customer-specific isolation as a platform – not a project." We protect data-in-use so you can safely run agents. If you are deploying AI in your company, we would love to talk to you. We will be at the Glean Go conference. cc Arvind Jain
-
Anshu Sharma reposted thisAnshu Sharma reposted thisWhat does it mean to be AI-native? At Glean:GO, Baseten, Enterpret, Fireworks AI, Gamma, Horizon AI, Seclore, Skyflow, and StackOne are answering that question, showing you how to turn bold AI ideas into working products, powerful experiences, and practical workflows for the enterprise. Catch the AI-Native Showcase during the morning keynote on day 2: https://lnkd.in/eqQAciPT #GleanGO
-
Anshu Sharma shared thisNext week, we are excited to be partnering with Glean for their conference in SF. The speakers will include my friends Jeetu Patel, Arvind Jain, and leaders from OpenAI and customers including our shared customers. I will be one of the speakers and looking forward to sharing how we put runtime data controls for secure AI, and sovereign AI. Our team will be onsite to answer your questions on how to safely deploy AI. See you there!
-
Anshu Sharma shared thisThank you IIT Kgp for the Distinguished Alumnus award, and teaching me the joy of just going for it. My time at IIT shaped me professionally, and personally. I learned from the hard core classes, from the labs, and from the professors. I also learned from my brilliant classmates who challenged me, competed with me - and collaborated in projects with me. While four years far from home - 20 hours by train - were tough, I loved being on the campus and spending time there. It was our little cocoon where you could unashamedly pursue whatever interests you. I spent most of my time pursuing the understanding of how computers work and how to make them do things - and that pursuit has continued through out my life. I love building software. It gives me joy. Thank you Anand and the nomination committee. Go IIT Kgp!
-
Anshu Sharma shared thisHow do you implement data controls for AI with the gateway you already use? Join Kong and Skyflow as we walk you through secure and sovereign AI. Register to join now or listen to recording.
-
Anshu Sharma shared thisIf you are not living under a rock, you must have read about OpenAI agents hacking into Hugging Face to unlock its goals, and Anthropic announcing similar challenges with its own models during testing. Combine this with the not-really-zero zero data retention for latest models. How do you operate AI in such an environment while meeting your obligations to your customers? When you are sending your customers' data to models, what security controls do you have in place? Join us as Skyflow and Kong partner to help lay out the architecture of data controls for agents & models. cc Augusto Marietti
-
Anshu Sharma reacted on thisAnshu Sharma reacted on thisIt has been a great experience to be part of a team that is technically awesome.. our team that built our Advertisement (Ads) technical stack. Ads is the reason we can provide ChatGPT for free to 1 billion users worldwide. In about 200 days, we scaled from $0 to $1 billion annualized revenue to be the fastest growing Ads business is history. Rockset (acquired by OpenAI) is the AI-search and realtime-analytical database that powers the Ads engine ...candidate generation via ANN search, product feed search, etc. https://lnkd.in/gCeqp4Xq
-
Anshu Sharma reacted on thisAnshu Sharma reacted on thisA year ago at Glean:GO, we announced #GleanProtect. At the time, the question was straightforward: “How do we help enterprises move quickly with AI while keeping their data and people safe?” Since then, AI has expanded far beyond a single assistant. It now lives across copilots, agents, coding tools, internal applications, MCP connections, and workflows that can read data, use tools, and take action. At this year’s Glean:GO, we shared how #GleanProtect has evolved with AI itself. It has become a security and governance layer for the broader AI ecosystem — across context, memory, skills, agents, models, and tools: 🧠 Using context-aware threat detection to distinguish legitimate agent workflows from risky activity such as data harvesting, exfiltration, or destructive writes 🔐 Securing sensitive and overshared data before it becomes an AI-powered exposure 💬 Helping security teams triage findings and turn large queues into prioritized action using Glean Assistant powered Security Chat 🧩 Governing the memory and skills that agents rely on—including the instructions, knowledge, and permissions that shape their behaviour 🤖 Applying policies across agents, users, tools, and data, with the precision to protect critical systems without blocking legitimate work 🌐 Extending Protect through the Glean AI Gateway, so teams can govern models, AI applications, and MCP tools through a consistent control plane 🛡️ Adding safeguards for risks such as prompt injection, malicious code, restricted topics, and unsafe browser activity for every interaction These updates have resonated deeply with our customers as they double down on AI adoption across their organisations this year. As I continue building #GleanProtect alongside Sunil Agrawal, Kiran Bondalapati, Allan Livingston, Harsh Singhal and the rest of our incredible team, our goal is to help them move quickly and scale confidently — without compromising on security. #GleanGO #GleanProtect #EnterpriseAI #AIGovernance #Cybersecurity
-
Anshu Sharma reacted on thisAnshu Sharma reacted on thisOur Q2 was about one thing, winning with AI. The results speak for themselves. https://bit.ly/4gkGax0 And while we can say this was our doing, the real heroes are our customers. Thank you. We understand the pressure and inherent contradictions you face every day. Roll out agents now, but do it securely. Democratize AI, but make sure its operating within guardrails. Oh, and make sure its not blowing up the token budget while you're at it. We're with you - listening, learning and building product. We're just getting started, and you have our unwavering commitment to make your agentic ambitions true - securely. #OktaSecuresAI Okta
Experience & Education
Volunteer Experience
Publications
Patents
-
Methods and Systems for Shared Data Sets in an On-Line Services Environment
Issued US US20130232165
Honors & Awards
-
TiE Charter Member
TiE Silicon Valley
Recommendations received
12 people have recommended Anshu
Join now to viewOther similar profiles
Explore more posts
-
Maria Palma
On-Prem is the New Cloud: Why Enterprises are Reversing Stance in the Age of AI Remember when cloud was the battleground for innovation, agility, and scale? Everyone was tracking the % of on-prem compute moving to the cloud and it was growing every year? Well now there’s a reversal. The reality today? On-prem is the new cloud. Here’s why. AI’s Dark Side: Legal Risk and Data Governance Headaches Enterprises are excited about AI, but also uneasy as vendors like OpenAI face intensifying legal pressure. In a high-stakes lawsuit, a judge ordered OpenAI to preserve all ChatGPT chats (even deleted ones), overriding its 30-day deletion policy. Imagine that: you delete sensitive conversations, but they’re still sucked into legal limbo. The precedent is clear: AI prompts and outputs are discoverable records, demanding integration into enterprise ESI (Electronic Stored Information) policies. For large organizations handling sensitive data, the risk is simple: you can’t afford your AI tools to turn into legal liabilities. Especially when global regulations (GDPR, etc.) and internal governance collide with unpredictably broad discovery orders. Sovereignty, Control, Trust Many enterprises still ban ChatGPT or impose strict rules on usage. Even when allowed, token limits are throttled. Enterprises are cautiously letting AI in (as they should). Enterprises need to have control and trust. SAP is leaning into this with its Sovereign Cloud On-Site, deployable on customer premises with SAP-managed infrastructure. It is built to deliver data, operational, technical, and legal sovereignty in one package. Doubt it will be the last. Scaling Pains Talking to founders, three reasons stand out why cloud isn’t scaling with AI: 1/ The rate at which AWS/GCP lets you provision compute and the compute it wants you to provision both falter as AI and AI driven developers deploy a lot more code more frequently than before. So sometimes it’s not about data governance, but actually about performance and functionality. 2/ Shadow AI is rampant—employees paste sensitive info into models despite restrictions—creating demand for better permissioning tools. 3/ Data leakage remains a live risk, like the August 2025 Grok incident where hundreds of thousands of private chats became publicly accessible and indexed. One thing is for sure - founders selling into the enterprise are scaling faster if they offer the ability to run on-prem. The Future is Hybrid Don’t get me wrong - cloud is not going anywhere. But neither is on prem. A few years agoI thought we were headed for a full cloud shift, but now I think the future is truly hybrid.
4 Comments -
Vishal Lugani
Bilt CEO/founder Ankur Jain published his annual letter today. Publicly posting an annual letter requires commitment to constantly iterating and continuously raising the bar. It's easy to forget how hard it is to build a breakout company, particularly one that pioneers a whole new category defining set of ideas. In this letter, he talks about the significant progress the business has made and is honest about the challenges that come with hyper growth. The company started with the belief that commerce starts where the home is. Shopify made it easy for merchants to sell to consumers online. Ecommerce penetration in the US is still <20% of total sales. Focused on the rest of the pie, Bilt has built the platform for neighborhood commerce. Consumers can earn rewards for rent and mortgage payments and all their neighborhood spend alike. And when they leave their neighborhood to travel, Bilt provides them the richest set of rewards options on the market. This year the company is on track to process $100B+ in housing payments. To boot, Bilt enables real estate owners and managers to offer renters a modern hospitality platform integrated into the systems that touch maintenance, amenities, move-in logistics and more. Neighborhood merchants also benefit from Bilt's software and are able to reward Bilt members who dine, ride, or spend with them. Bilt members benefit from the company's new AI concierge and can earn rewards whether they have a Bilt co-brand credit card or not. On the topic of their credit card, as the company transitioned from their prior one-card-only offering to multiple card options (e.g., Bilt 2.0), they saw 83% of original card holders request one of the Bilt 2.0 cards. 30%+ of that group opted into premium cards. And since that Bilt 2.0 transition a few months ago, cardholders are spending 2.1x what they previously were on non-housing spend. Neighborhood partners who work with Bilt are seeing significant uplift in Bilt member spend. Onwards to the team (cc: David M. Wyler, Daniel Seder, Brandt Smallwood, Zoe Oz) and company. Bilt is the membership for where you live.
2 Comments -
Larry Cheng
Good piece for Entrepreneur by my Volition Capital colleague, Tiger Henderson. We often see companies that have built their business on a single platform like Amazon, Shopify, Salesforce. Tiger walks through the key questions we ask on evaluating that risk: 1. Are incentives aligned with the underlying platform? 2. Who owns the end-user relationship? 3. Does the product embed itself deeply in workflows, or just sit on a feature layer? 4. Does the underlying platform have a history of developing its 3rd party ecosystem? 5. Does the CEO have a defined strategy to derisk the business over time? You can read the full article here: https://lnkd.in/epWhpcsd
3 Comments
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content