Ashutosh Kulkarni’s Post

Search powered AI brought to you by Elastic, listen to our CEO pov, this is here 👇

What we are doing technically at elastic to enable enterprises realize the value of data by providing a robust data platform which is pretty awesome! Our tech allows gen ai engines to provide context specific results for your enterprise easily. Being a techie at heart, I love the scale capabilities we bring to our customers so that they can run their businesses effectively #genai #elasticsearch #elastic

Innovation matters! Let’s have a talk how you can use it for your business.

📢 Too many businesses are obsessed with picking the perfect GenAI model, missing out on other crucial aspects. At Aviso, we believe that choosing a model is just the beginning. Integrating generative AI like GPT-4o, Google’s Gemini, and more involves unique challenges—costs, integration hurdles, data privacy risks, and more. That’s why Aviso offers tailored solutions to ensure seamless integration, performance optimization, and data security. Here are the top 5 factors that set our GenAI approach different from the lot: 🔹 Blend of open-source and proprietary models 🔹 Cutting-edge optimization strategies for RAG and fine-tuning 🔹 Robust guardrails to mitigate risks like model lock-in 🔹 Expertise in selecting the right vector databases 🔹 Platform-agnostic for seamless integration And there's more to it. Check out our blog to learn more about what makes Aviso's GenAI approach unique: https://lnkd.in/g5Rba3RD #GenerativeAI #AIForsales #DataPrivacy #AvisoAI

Today I want to share an interesting project that is working on ZeroKnowledge proofs of data that will be stored on the L1 settlement layer. The metadata itself will ultimately be tied to its underlying dataset, as these datasets are stored themselves on a data ledger. Grass provides the data layer for a modular AI stack, and these developments will lay the groundwork for greater transparency and rewards for node providers that are proportionate to the amount of work they perform. Grass: The First Ever Layer 2 Data Rollup Get your grass 🤔 here - get paid for unused internet. https://lnkd.in/exY9kqMC How a Layer Two Will Establish Data Provenance  "The world needs a method for proving the origin of AI training data, and that’s what Grass is now building. Soon, every time data is scraped by Grass nodes, metadata will be recorded to verify the website it was scraped from. This metadata will then be permanently embedded in every dataset, enabling builders to know its source with total certainty. They can then share this lineage with their users, who can rest easier knowing that the AI models they interact with were not deliberately trained to give misleading answers.   This will be a big lift and involve a major expansion of our protocol as we prepare for scraping operations to reach tens of millions of web requests per minute. Each of these will need to be validated, which will take more throughput than any L1 can provide. That’s why we’re announcing our plan to build a layer 2 solution to handle this significant upgrade to our capabilities. The L2 will be a sovereign rollup, featuring a ZK processor so that metadata can be batched for validation and used to provide a persistent lineage for every dataset we produce. This is what it will take for the base layer of all AI development to advance to the next stage." https://lnkd.in/eZU6BjGM Let me know your thoughts, experience or other ideas making the world a better and safer place.

There is currently no standard for Test and Evaluation (T&E) in AI, but it is a critical component of developing responsible, trustworthy models. That's why today - alongside our collaboration with The White House and DEF CON in an evaluation of the leading models - Scale is releasing its T&E methodology and platform in early access, to help model builders deploy LLMs safely and securely. Scale’s role as the data partner for players across the AI ecosystem (including OpenAI, Meta, Microsoft, and Anthropic) has uniquely allowed us to develop an approach to effectively test and evaluate LLMs, and we are excited to be releasing our methodology freely for the industry to learn from. Our first T&E offering for LLMs builds upon this approach and helps model builders evaluate their LLMs capabilities and vulnerabilities. Our T&E platform is already being leveraged by OpenAI for red-teaming their models, and now select customers at the forefront of model development can request early access. For AI to reach its full potential, it’s critical the industry is working to address these known challenges in parallel with other model advancements. Learn more about our T&E methodology and platform here: 

🚀 Exciting News for AI Enthusiasts! 🚀 At Datadog's DASH 2024, we unveiled some groundbreaking updates, and I'm thrilled to share one that stands out for all teams working with generative AI: LLM Observability. 🌟 What is LLM Observability? It's a powerful new tool designed to give you deep visibility into your AI applications. Working with models like GPT, Claude, and Gemini? This is for you. Here's what LLM Observability can do: 🔍 Trace AI Applications: Diagnose errors across every component, evaluate performance, and identify drifts in topics and responses. 🔐 Enhance Security: Detect and prevent prompt injections and PII leakage. ⚡ Optimize Performance: Identify and resolve latency issues to keep your applications running smoothly. Why should you care? LLM Observability helps you: - Quickly diagnose and fix issues without extensive manual searches. - Ensure your AI models meet quality and performance standards. - Maintain data security and comply with privacy regulations. LLM Observability is now available and can be seamlessly integrated into your Datadog account. Interested in learning more? Drop me a message, and let's discuss how this can benefit your team and take your AI applications to the next level! #AI #MachineLearning #GenerativeAI #Observability #DataSecurity #Datadog #TechInnovation

A well-designed data strategy is key to any modern business. With global data/information growing exponentially, it's crucial to leverage analytics, AI, and ML technologies to stay ahead. In upcoming posts, I'll delve into key data considerations businesses must prioritize for growth, including data cleanliness, business requirements, integrations, analytics and insights, security and privacy, monitoring, and AI/machine learning. Stay tuned! #DataStrategy #BusinessGrowth #Analytics #AI #MachineLearning #DataPrivacy #DataSecurity #DataInsights (Statista, 2024)