NewGet @lindavivah's full report
instagram analytics
Similar Accounts:
followers
116K
impressions
10.6M
likes
142K
comments
13.7K
posts
148
engagement
1.41%
emv
$266K
Average per post
71.9K

Key Metrics

Distributions

Top Content

My AI agent caught a medical bill I’d missed. I gave my personal life a new operating system and it took minutes to set up. Meet Pearl, my personal agent in the new @muse app from @meta (yes, you name your own). #metapartner Time to get my ducks in a row for back to school season... but really it’s 40 things living in my head 🦆🫠 School emails, family pickups, appointments, bills, birthday parties and hosting, gifts, trying to get myself back to the gym, plus running my business on top of it all. Pearl is proactive, she does things in advance. That’s how she caught the bill: a daily audit of my inboxes... she lined it up and brought me straight to the approval. She makes everything as easy as it can be while I keep the final yes. Because of the connectors she has real context from day one, and over time she’ll suggest things on her own based on what she’s seeing. As someone who builds agentic systems I was super impressed by how intuitive it all is. Under the hood, here’s how it works: 🔹 Every muse is a secure computer in the cloud (a VM) with its a browser you share... so Pearl keeps working even when I close the app 🔹 A separate Sentinel agent has to approve anything before it leaves that machine 🔹 She never sees my passwords or card: credentials sit in secure storage she can use but can’t read (like a valet key) 🔹 Checkout generates a one-time card so my real card stays hidden from the merchant 🔹 I choose what she connects to and there’s an audit trail of everything she’s done and plans to do She’s also got my gym goal, my morning brief and a watch list... and if a task needs a tool that doesn’t exist I just ask her to build it. The code’s right there, I can open up exactly what she built. Muse is free to start... try the new Muse app from Meta & curious what you named yours? 👇 Businesses and products shown are AI-generated suggestions, not sponsored, endorsed by, or affiliated with Meta.
5.22M
7.86K
160
1mo ago
lindavivah
My AI agent caught a medical bill I’d missed. I gave my personal life a new operating system and it took minutes to set up. Meet Pearl, my personal agent in the new @muse app from @meta (yes, you name your own). #metapartner Time to get my ducks in a row for back to school season... but really it’s 40 things living in my head 🦆🫠 School emails, family pickups, appointments, bills, birthday parties and hosting, gifts, trying to get myself back to the gym, plus running my business on top of it all. Pearl is proactive, she does things in advance. That’s how she caught the bill: a daily audit of my inboxes... she lined it up and brought me straight to the approval. She makes everything as easy as it can be while I keep the final yes. Because of the connectors she has real context from day one, and over time she’ll suggest things on her own based on what she’s seeing. As someone who builds agentic systems I was super impressed by how intuitive it all is. Under the hood, here’s how it works: 🔹 Every muse is a secure computer in the cloud (a VM) with its a browser you share... so Pearl keeps working even when I close the app 🔹 A separate Sentinel agent has to approve anything before it leaves that machine 🔹 She never sees my passwords or card: credentials sit in secure storage she can use but can’t read (like a valet key) 🔹 Checkout generates a one-time card so my real card stays hidden from the merchant 🔹 I choose what she connects to and there’s an audit trail of everything she’s done and plans to do She’s also got my gym goal, my morning brief and a watch list... and if a task needs a tool that doesn’t exist I just ask her to build it. The code’s right there, I can open up exactly what she built. Muse is free to start... try the new Muse app from Meta & curious what you named yours? 👇 Businesses and products shown are AI-generated suggestions, not sponsored, endorsed by, or affiliated with Meta.
👀 NEW Claude can now control your computer: Mouse, keyboard, and screen, giving it the ability to use any app without an MCP or connector. It does first check if you have a connector an will prompt you for permission. Read the “Claude Cowork Safety”official doc if you plan to use this. Here I’m using dispatch via Claude cowork to search and export one of the videos I edited locally in CapCut 🤯 (It did prompt me for specific permissions) It can open your apps, navigate your browser, fills in spreadsheets….anything you’d do sitting at your desk. Currently works on Mac for both Claude Cowork & Claude Code [ai engineer, tech news, agentic ai]
1.02M
23.9K
256
7mo ago
lindavivah
👀 NEW Claude can now control your computer: Mouse, keyboard, and screen, giving it the ability to use any app without an MCP or connector. It does first check if you have a connector an will prompt you for permission. Read the “Claude Cowork Safety”official doc if you plan to use this. Here I’m using dispatch via Claude cowork to search and export one of the videos I edited locally in CapCut 🤯 (It did prompt me for specific permissions) It can open your apps, navigate your browser, fills in spreadsheets….anything you’d do sitting at your desk. Currently works on Mac for both Claude Cowork & Claude Code [ai engineer, tech news, agentic ai]
Context vs Memory vs Harness engineering explained in 40 seconds by Richmond Alake ⚡️ 💡These 3 disciplines are core to building agents that can actually remember and reason over time Most agents work well within a single session but lose everything the moment it ends. Memory engineering treats long-term memory as first-class infrastructure. Richmond Alake (Director of AI DevEx at Oracle) walked me through all three in NYC🗽 ✈️ I’m PSYCHED to be heading to an offsite with the Oracle developer team very soon to go even deeper on agent harnesses & excited to share learnings with y’all! If you want to learn how to build memory-aware agents, Richmond has a free short DeepLearning.AI course in collaboration with Oracle that he co-taught with his colleague Nacho! 📚 Comment ✨DEVELOPER✨ and I’ll DM you the link to the free DeepLearning agent memory course & the AI Developer GitHub repo by Oracle to help you start building 💻
818K
29.0K
1.40K
6mo ago
lindavivah
Context vs Memory vs Harness engineering explained in 40 seconds by Richmond Alake ⚡️ 💡These 3 disciplines are core to building agents that can actually remember and reason over time Most agents work well within a single session but lose everything the moment it ends. Memory engineering treats long-term memory as first-class infrastructure. Richmond Alake (Director of AI DevEx at Oracle) walked me through all three in NYC🗽 ✈️ I’m PSYCHED to be heading to an offsite with the Oracle developer team very soon to go even deeper on agent harnesses & excited to share learnings with y’all! If you want to learn how to build memory-aware agents, Richmond has a free short DeepLearning.AI course in collaboration with Oracle that he co-taught with his colleague Nacho! 📚 Comment ✨DEVELOPER✨ and I’ll DM you the link to the free DeepLearning agent memory course & the AI Developer GitHub repo by Oracle to help you start building 💻
Why do the best ideas always hit when you are away from your desk 🌱 I was in Central Park and was able to tag Devin (@trycognition ) in Slack with my prompt to built AgentGeoScore, a site where you can check if your site is optimized for agent search Comment ✨DEVIN✨ & I’ll DM you the link to the GitHub repo & Devin to try it out for yourself! I also open sourced the code for agentgeoscore & will be maintaining it so feel free to contribute! I was super impressed but Devin’s native slack integration and how so many features are built-in and enterprise grade + easy to use out of the box. I was able to build so much just in the slack thread and get everything from screen recordings of the site, automatic security scans, merge from a button in the thread, Devin deploy right in the thread to see the site easily from anywhere & so much more! & since I used it via slack on my phone, context is never lost. If you try it out would love to hear your thoughts! & as always happy building!
553K
323
55
5mo ago
lindavivah
Why do the best ideas always hit when you are away from your desk 🌱 I was in Central Park and was able to tag Devin (@trycognition ) in Slack with my prompt to built AgentGeoScore, a site where you can check if your site is optimized for agent search Comment ✨DEVIN✨ & I’ll DM you the link to the GitHub repo & Devin to try it out for yourself! I also open sourced the code for agentgeoscore & will be maintaining it so feel free to contribute! I was super impressed but Devin’s native slack integration and how so many features are built-in and enterprise grade + easy to use out of the box. I was able to build so much just in the slack thread and get everything from screen recordings of the site, automatic security scans, merge from a button in the thread, Devin deploy right in the thread to see the site easily from anywhere & so much more! & since I used it via slack on my phone, context is never lost. If you try it out would love to hear your thoughts! & as always happy building!
7 core Claude Code features explained in 90 seconds.. let’s go! ⚡ To unlock Claude Code’s full potential you need to know what’s available to you. Here’s the TLDR: 🔷 CLAUDE.md → a markdown file Claude loads as context each session 🔷 Skills → packaged instructions Claude loads when the task matches 🔷 MCP → standard protocol for connecting Claude to external tools and data 🔷 Hooks → commands that run automatically at points in Claude’s lifecycle 🔷 Subagents → isolated Claude instances with their own context and tool access 🔷 Plugins → skills, hooks, and MCP configs bundled into one installable unit 🔷 Agent Teams → parallel Claude sessions with a lead, messaging each other directly
293K
9.48K
171
6mo ago
lindavivah
7 core Claude Code features explained in 90 seconds.. let’s go! ⚡ To unlock Claude Code’s full potential you need to know what’s available to you. Here’s the TLDR: 🔷 CLAUDE.md → a markdown file Claude loads as context each session 🔷 Skills → packaged instructions Claude loads when the task matches 🔷 MCP → standard protocol for connecting Claude to external tools and data 🔷 Hooks → commands that run automatically at points in Claude’s lifecycle 🔷 Subagents → isolated Claude instances with their own context and tool access 🔷 Plugins → skills, hooks, and MCP configs bundled into one installable unit 🔷 Agent Teams → parallel Claude sessions with a lead, messaging each other directly
Context Engine Architecture explained in 2 min ⏱️ .. walk with us in SF! Love how the Head of Context Engine @ Redis, Simba Khadder explains why we need this architecture: “Now we have agents that have hands. They need to be able to get data, act on data, search for more data.” Here’s the Context Engine Architecture breakdown: 🔹 Systems of record: your CRMs, databases, all the places your data already lives (you don’t want to wire an agent into 55 of them) 🔹 ETL: pulls it all into one materialized view, basically one live copy of your data that agents can actually use 🔹 Semantic layer: describes that data the way you’d explain it to a person: orders, users, items, not fields and tables 🔹 MCP: the open standard that lets the agent go grab that context on its own 🔹 Internal context: the agent also remembers what it’s done and caches answers it already figured out, so it doesn’t solve the same thing twice 💡 Put together, context stays fresh, navigable, and fast, and it compounds as the agent works. That’s the general pattern, whatever stack you use. Redis ships it as Redis Iris: RDI, Context Retriever, Agent Memory, and LangCache. Comment ✨CONTEXT✨ & I’ll send you the link to try Redis Iris for yourself for free
149K
3.09K
244
2mo ago
lindavivah
Context Engine Architecture explained in 2 min ⏱️ .. walk with us in SF! Love how the Head of Context Engine @ Redis, Simba Khadder explains why we need this architecture: “Now we have agents that have hands. They need to be able to get data, act on data, search for more data.” Here’s the Context Engine Architecture breakdown: 🔹 Systems of record: your CRMs, databases, all the places your data already lives (you don’t want to wire an agent into 55 of them) 🔹 ETL: pulls it all into one materialized view, basically one live copy of your data that agents can actually use 🔹 Semantic layer: describes that data the way you’d explain it to a person: orders, users, items, not fields and tables 🔹 MCP: the open standard that lets the agent go grab that context on its own 🔹 Internal context: the agent also remembers what it’s done and caches answers it already figured out, so it doesn’t solve the same thing twice 💡 Put together, context stays fresh, navigable, and fast, and it compounds as the agent works. That’s the general pattern, whatever stack you use. Redis ships it as Redis Iris: RDI, Context Retriever, Agent Memory, and LangCache. Comment ✨CONTEXT✨ & I’ll send you the link to try Redis Iris for yourself for free
Principal AI Architect & Engineer @codewithbrij walks through the 8 layers from LLM ➡️ Agentic AI Summary of the stack: 🧠 LLM → predicts the next word. Smart, but frozen. 📚 + RAG → reads your docs before answering. 🔧 + Tool Calling → now it can DO things. Call APIs, run code, query DBs. 💾 + Memory → remembers yesterday. Learns what works. 🤖 = AI Agent → LLM + RAG + Tools + Memory. 🤝 + Agentic AI → agents delegating to agents. Planning. Self-correcting. 🧩 + Skills & Hooks → modular know-how + lifecycle triggers. 🛡️ + Governance & Observability → traces, guardrails, audit. This is what makes it production. It was SO great getting to meet in person a few weeks back at the Oracle dev offsite! Check out Brij’s page @codewithbrij for the best technical infographics + AI education! And he is also one of the nicest people!
140K
3.51K
41
4mo ago
lindavivah
Principal AI Architect & Engineer @codewithbrij walks through the 8 layers from LLM ➡️ Agentic AI Summary of the stack: 🧠 LLM → predicts the next word. Smart, but frozen. 📚 + RAG → reads your docs before answering. 🔧 + Tool Calling → now it can DO things. Call APIs, run code, query DBs. 💾 + Memory → remembers yesterday. Learns what works. 🤖 = AI Agent → LLM + RAG + Tools + Memory. 🤝 + Agentic AI → agents delegating to agents. Planning. Self-correcting. 🧩 + Skills & Hooks → modular know-how + lifecycle triggers. 🛡️ + Governance & Observability → traces, guardrails, audit. This is what makes it production. It was SO great getting to meet in person a few weeks back at the Oracle dev offsite! Check out Brij’s page @codewithbrij for the best technical infographics + AI education! And he is also one of the nicest people!
Andrew Ng & the @deeplearningai team analyzed 10,000+ job postings + expert interviews & just shared the skills that matter most in AI engineering right now for developers 📚 The 4 skills: 🔹 Building & deploying AI applications: AI systems have unpredictable outputs, so the work is learning the building blocks (RAG, evals, agentic workflows, context engineering) and using them to measure, steer, and govern how the system behaves 🔹 Software engineering fundamentals: so much of software engineering is actually trade-offs (cost, speed, scalability, reliability, security). It’s making decisions, and if you don’t know the trade-offs, it’s hard to make those decisions 🔹 Using coding agents: building a mental model of what agents can and can’t do, managing context, balancing planning vs execution, and keeping current as the tools evolve 🔹 Shaping the build: participating in the spec decisions themselves, which takes product sense and an understanding of business context and customer goals. Part of the skill is knowing when to spin up an MVP and when to build carefully Comment SKILLS and I’ll send you the link to the report 📩
136K
3.47K
775
2mo ago
lindavivah
Andrew Ng & the @deeplearningai team analyzed 10,000+ job postings + expert interviews & just shared the skills that matter most in AI engineering right now for developers 📚 The 4 skills: 🔹 Building & deploying AI applications: AI systems have unpredictable outputs, so the work is learning the building blocks (RAG, evals, agentic workflows, context engineering) and using them to measure, steer, and govern how the system behaves 🔹 Software engineering fundamentals: so much of software engineering is actually trade-offs (cost, speed, scalability, reliability, security). It’s making decisions, and if you don’t know the trade-offs, it’s hard to make those decisions 🔹 Using coding agents: building a mental model of what agents can and can’t do, managing context, balancing planning vs execution, and keeping current as the tools evolve 🔹 Shaping the build: participating in the spec decisions themselves, which takes product sense and an understanding of business context and customer goals. Part of the skill is knowing when to spin up an MVP and when to build carefully Comment SKILLS and I’ll send you the link to the report 📩
🎮NVIDIA & Microsoft brought a hands-on arcade for AI developers to Microsoft Build, & it’s available on-demand! If you didn’t get a chance to be in the room, this is a fantastic self-paced way to go from your first inference all the way to production-ready deployment. Comment ✨ARCADE✨& ill DM you the link! @nvidia @microsoft
115K
2.77K
1.51K
4mo ago
lindavivah
🎮NVIDIA & Microsoft brought a hands-on arcade for AI developers to Microsoft Build, & it’s available on-demand! If you didn’t get a chance to be in the room, this is a fantastic self-paced way to go from your first inference all the way to production-ready deployment. Comment ✨ARCADE✨& ill DM you the link! @nvidia @microsoft
🤩 Claude Code just launched Channels! Let’s set it up together 🎉 Channels allow you to control your Claude Code sessions through select MCP’s, starting with Telegram and Discord Comment ✨CHANNELS✨ & i’ll send the doc to your DM [AI Engineer, agentic AI, coding]
99.6K
2.73K
530
7mo ago
lindavivah
🤩 Claude Code just launched Channels! Let’s set it up together 🎉 Channels allow you to control your Claude Code sessions through select MCP’s, starting with Telegram and Discord Comment ✨CHANNELS✨ & i’ll send the doc to your DM [AI Engineer, agentic AI, coding]
Traditional RAG vs Graph RAG explained in 60 seconds by Nacho Martínez ⚡ 💡 Both pull external knowledge into an LLM. The difference is whether what comes back is isolated, or actually connected. 🔷 Traditional RAG chunks your docs, vectorizes each chunk, similarity search returns the top matches. Each piece stands alone. 🔷 Graph RAG organizes knowledge as a graph with edges between related items. Retrieval can follow those edges, so you get a match plus what it connects to. Some implementations store chunks as nodes and pull in neighbors. Others build a graph of entities and their relationships pulled from your docs. Same idea, different units. 🏢 i recently got to visit the Oracle office and spend the day with their amazing devrel team going deep on the Oracle AI Database, learning about their agentic AI work, and doing hands-on workshops. SO much good stuff. One thing that genuinely blew me away was that the Oracle AI database has JSON Relational Duality. You store your data once and access it as both relational rows AND JSON documents. No sync, no duplication 🔥 📚 Nacho Martínez (Data Scienctist Advocate at Oracle) shared a fantastic open source AI Developer Hub the team’s been building, includes notebooks on agentic RAG with hybrid search, RAG with evals, RAG zero-to-hero, and more. 💻 Comment ✨DEVELOPER✨ and I’ll DM you the link to repo to start building
95.5K
3.57K
70
5mo ago
lindavivah
Traditional RAG vs Graph RAG explained in 60 seconds by Nacho Martínez ⚡ 💡 Both pull external knowledge into an LLM. The difference is whether what comes back is isolated, or actually connected. 🔷 Traditional RAG chunks your docs, vectorizes each chunk, similarity search returns the top matches. Each piece stands alone. 🔷 Graph RAG organizes knowledge as a graph with edges between related items. Retrieval can follow those edges, so you get a match plus what it connects to. Some implementations store chunks as nodes and pull in neighbors. Others build a graph of entities and their relationships pulled from your docs. Same idea, different units. 🏢 i recently got to visit the Oracle office and spend the day with their amazing devrel team going deep on the Oracle AI Database, learning about their agentic AI work, and doing hands-on workshops. SO much good stuff. One thing that genuinely blew me away was that the Oracle AI database has JSON Relational Duality. You store your data once and access it as both relational rows AND JSON documents. No sync, no duplication 🔥 📚 Nacho Martínez (Data Scienctist Advocate at Oracle) shared a fantastic open source AI Developer Hub the team’s been building, includes notebooks on agentic RAG with hybrid search, RAG with evals, RAG zero-to-hero, and more. 💻 Comment ✨DEVELOPER✨ and I’ll DM you the link to repo to start building
Love the way @gregisenberg phrased it 💡 Comment ✨OPTIMIZE✨ and I’ll send you 5 prompts for your vibe coding tool of choice to help your site get read and cited by AI agents 🌐
75.6K
2.39K
1.68K
6mo ago
lindavivah
Love the way @gregisenberg phrased it 💡 Comment ✨OPTIMIZE✨ and I’ll send you 5 prompts for your vibe coding tool of choice to help your site get read and cited by AI agents 🌐
There has been an 800% increase in FDE (forward deployed engineer) job postings in the past year 🤯 So I had to ask the INCREDIBLE Leah McGowen-Hare, SVP of Forward Deployed Engineering @salesforce , whose team tripled in just six months, how they define the role and what they look for when they hire Last month I attended Agentforce World Tour NY and Leah’s keynote was absolutely fantastic! One of the best speakers I’ve seen. I was fangirling 🥹🤍 An FDE is a deep technical expert across the Agentforce suite, part tech guru, part business consultant, the person who can make or break an AI agent launch. Leah’s take: they’re the unicorns on the front line with customers, with a direct line back to product so what they learn in the field shapes the roadmap. And who they hire for it? The curious ones. The constant learners and tinkerers who want to stay on the brink of the newest tech #SalesforcePartner
74.8K
1.50K
51
5mo ago
lindavivah
There has been an 800% increase in FDE (forward deployed engineer) job postings in the past year 🤯 So I had to ask the INCREDIBLE Leah McGowen-Hare, SVP of Forward Deployed Engineering @salesforce , whose team tripled in just six months, how they define the role and what they look for when they hire Last month I attended Agentforce World Tour NY and Leah’s keynote was absolutely fantastic! One of the best speakers I’ve seen. I was fangirling 🥹🤍 An FDE is a deep technical expert across the Agentforce suite, part tech guru, part business consultant, the person who can make or break an AI agent launch. Leah’s take: they’re the unicorns on the front line with customers, with a direct line back to product so what they learn in the field shapes the roadmap. And who they hire for it? The curious ones. The constant learners and tinkerers who want to stay on the brink of the newest tech #SalesforcePartner
Let’s see if I can cover the ML pipeline in 60 seconds ⏰😅 Machine learning isn’t just training a model. A production ML lifecycle typically looks like this: 1️⃣ Define the problem & objective 2️⃣ Collect and (if needed) label data 3️⃣ Split into train / validation / test sets 4️⃣ Data preprocessing & feature engineering 5️⃣ Train the model (forward pass + backpropagation in deep learning) 6️⃣ Evaluate on held-out data to measure generalization 7️⃣ Hyperparameter tuning (learning rate, architecture, etc.) 8️⃣ Final testing before release 9️⃣ Deploy (batch inference or real-time serving behind an API) 🔟 Monitor for data drift, concept drift, latency, cost, and reliability 1️⃣1️⃣ Retrain when performance degrades Training updates weights. Evaluation measures performance. Deployment serves predictions. Monitoring keeps the system healthy. It’s not linear. It’s a loop. And once you move beyond a single experiment, that loop becomes a systems problem. At scale, the challenge isn’t just modeling … it’s building reliable, scalable infrastructure that supports the entire lifecycle. Curious if this type of content is helpful! Lmk in the comments & as always Happy Building! 🤍
64.4K
3.81K
60
8mo ago
lindavivah
Let’s see if I can cover the ML pipeline in 60 seconds ⏰😅 Machine learning isn’t just training a model. A production ML lifecycle typically looks like this: 1️⃣ Define the problem & objective 2️⃣ Collect and (if needed) label data 3️⃣ Split into train / validation / test sets 4️⃣ Data preprocessing & feature engineering 5️⃣ Train the model (forward pass + backpropagation in deep learning) 6️⃣ Evaluate on held-out data to measure generalization 7️⃣ Hyperparameter tuning (learning rate, architecture, etc.) 8️⃣ Final testing before release 9️⃣ Deploy (batch inference or real-time serving behind an API) 🔟 Monitor for data drift, concept drift, latency, cost, and reliability 1️⃣1️⃣ Retrain when performance degrades Training updates weights. Evaluation measures performance. Deployment serves predictions. Monitoring keeps the system healthy. It’s not linear. It’s a loop. And once you move beyond a single experiment, that loop becomes a systems problem. At scale, the challenge isn’t just modeling … it’s building reliable, scalable infrastructure that supports the entire lifecycle. Curious if this type of content is helpful! Lmk in the comments & as always Happy Building! 🤍
Let’s build your own AI inference engine in 6 steps ⚙️ Why run your own? When you call an AI API, this whole layer runs for you, on someone else’s servers. Running it yourself gives you cost, privacy, and control 🧠 The full walkthrough is a free GitHub repo created by the Akamai Developers team. Comment ✨BUILD✨ & I’ll DM you where to find the repo and the Discord! This repo walks you through: 🔹 Spinning up a GPU (about 50 cents an hour) and deploying vLLM, an open-source inference server, with Qwen, a real open-weight model 🔹 Pointing your existing code at it (no rewrite needed) 🔹 And by doing this you learn what’s under the hood of AI inference, watching the KV cache (the model’s working memory) fill up live. That layer is a really important skill for engineers right now 🔹 Want a community of builders around it, or have questions as you go? That’s what Edge Case Discord is for: office hours, live build sessions with their engineers, and $300 in Akamai Cloud credits for new users Shout out to the Akamai team for awesome dev resources! Happy building! 🤍
62.6K
2.17K
1.14K
2w ago
lindavivah
Let’s build your own AI inference engine in 6 steps ⚙️ Why run your own? When you call an AI API, this whole layer runs for you, on someone else’s servers. Running it yourself gives you cost, privacy, and control 🧠 The full walkthrough is a free GitHub repo created by the Akamai Developers team. Comment ✨BUILD✨ & I’ll DM you where to find the repo and the Discord! This repo walks you through: 🔹 Spinning up a GPU (about 50 cents an hour) and deploying vLLM, an open-source inference server, with Qwen, a real open-weight model 🔹 Pointing your existing code at it (no rewrite needed) 🔹 And by doing this you learn what’s under the hood of AI inference, watching the KV cache (the model’s working memory) fill up live. That layer is a really important skill for engineers right now 🔹 Want a community of builders around it, or have questions as you go? That’s what Edge Case Discord is for: office hours, live build sessions with their engineers, and $300 in Akamai Cloud credits for new users Shout out to the Akamai team for awesome dev resources! Happy building! 🤍
Comment “MCP” & I’II send you the links via DM ✨ In late 2024, Anthropic introduced MCP (Model Context Protocol) because LLMs needed a cleaner, safer way to interact with the outside world. In just a year it already has over 97 million SDK downloads and 2 weeks ago Anthropic announced it’s donating MCP to the Linux Foundation. MCP is like giving models a shared language to access tools, trigger workflows, and pull data. The entire point of MCP is to make LLM’s more capable in a standardized way. These are some of the resources I found helpful when starting to build MCP servers & clients. I limited it to beginner friendly resources I thought were great but Imk if it would be helpful to share more advanced finds as well. Hope this helps others! Are there others you would add to the list? Lmk in the comments & as always Happy Building! Relevant tags 🏷️ [ai engineer, genai career, software engineer, software developer, MCP roadmap, how to learn ai, python for ai, lim from scratch, build in public, ai agent builder, deep learning 2025, ai portfolio project, ml roadmap, ai learning journey, agentic ai]
58.9K
1.31K
1.04K
10mo ago
lindavivah
Comment “MCP” & I’II send you the links via DM ✨ In late 2024, Anthropic introduced MCP (Model Context Protocol) because LLMs needed a cleaner, safer way to interact with the outside world. In just a year it already has over 97 million SDK downloads and 2 weeks ago Anthropic announced it’s donating MCP to the Linux Foundation. MCP is like giving models a shared language to access tools, trigger workflows, and pull data. The entire point of MCP is to make LLM’s more capable in a standardized way. These are some of the resources I found helpful when starting to build MCP servers & clients. I limited it to beginner friendly resources I thought were great but Imk if it would be helpful to share more advanced finds as well. Hope this helps others! Are there others you would add to the list? Lmk in the comments & as always Happy Building! Relevant tags 🏷️ [ai engineer, genai career, software engineer, software developer, MCP roadmap, how to learn ai, python for ai, lim from scratch, build in public, ai agent builder, deep learning 2025, ai portfolio project, ml roadmap, ai learning journey, agentic ai]
AI Engineer vs ML Engineer explained (while my youngest naps in Central Park 😂👶🍃) 🧠 ML engineers primarily focus on model training and performance optimization. That typically includes: • Data preprocessing and feature engineering • Designing and maintaining training pipelines • Selecting architectures and loss functions • Running experiments and tracking metrics • Hyperparameter tuning • Evaluating generalization performance • Scaling distributed training workloads Their center of gravity is improving how a model is trained and how well it performs. 🏗️ AI engineers primarily focus on system design and production deployment of AI capabilities. That typically includes: • Integrating trained or foundation models into applications • Designing RAG pipelines and agent architectures • Orchestrating tools, APIs, and external services • Managing state, retries, and failure handling • Implementing guardrails and evaluation frameworks • Optimizing latency, throughput, and cost • Scaling inference and serving infrastructure Their center of gravity is ensuring the AI system behaves reliably, safely, and efficiently in real-world environments. 🎯 Same end goal: production-ready AI. But they operate at different layers of the stack. 💡If you want a sticky way to remember it: ML engineers build and tune the brain. AI engineers build the nervous system and body around it. One optimizes how intelligence is trained. The other optimizes how intelligence is expressed and delivered. 🏷️ #AIEngineer #MLEngineer #DistributedSystems #LLMs #AgenticAI AIInfrastructure MachineLearning
51.5K
2.24K
77
8mo ago
lindavivah
AI Engineer vs ML Engineer explained (while my youngest naps in Central Park 😂👶🍃) 🧠 ML engineers primarily focus on model training and performance optimization. That typically includes: • Data preprocessing and feature engineering • Designing and maintaining training pipelines • Selecting architectures and loss functions • Running experiments and tracking metrics • Hyperparameter tuning • Evaluating generalization performance • Scaling distributed training workloads Their center of gravity is improving how a model is trained and how well it performs. 🏗️ AI engineers primarily focus on system design and production deployment of AI capabilities. That typically includes: • Integrating trained or foundation models into applications • Designing RAG pipelines and agent architectures • Orchestrating tools, APIs, and external services • Managing state, retries, and failure handling • Implementing guardrails and evaluation frameworks • Optimizing latency, throughput, and cost • Scaling inference and serving infrastructure Their center of gravity is ensuring the AI system behaves reliably, safely, and efficiently in real-world environments. 🎯 Same end goal: production-ready AI. But they operate at different layers of the stack. 💡If you want a sticky way to remember it: ML engineers build and tune the brain. AI engineers build the nervous system and body around it. One optimizes how intelligence is trained. The other optimizes how intelligence is expressed and delivered. 🏷️ #AIEngineer #MLEngineer #DistributedSystems #LLMs #AgenticAI AIInfrastructure MachineLearning
The beauty about tech is there is no one path. 🌱 One of the biggest lessons throughout my journey has been that what felt like the hardest parts often involved “figuring out my next growth step”. The decision is sometimes harder than the execution. ⚙️ Overtime, I learned those phases of “going in circles” are totally normal & actually a lot of the work within your journey. 🔁 For example, it took me a while to figure out that I would like to go in to SRE from webdev etc. 🔁 the same thing happened when I decided it was time to go to a startup and the AI infra end. Have you experienced something similar? Lmk in comments 🤎 ⬇️ 💡 Remember, YOU are the artist of your own life 🎨 Summary of my Untraditional Journey in to Tech & Cloud & AI: 💻 2013: graduated BA in Philosophy & started working in media 💻 2015: started teaching myself how to code & a few months later attended a coding bootcamp (@flatironschool Fullstack Web Dev immersive) 💻 2016: started working as a javaScript Web Developer in media 💻 2018: Started teaching myself Cloud Computing via certs & projects, specifically AWS 💻 2020: Shifted in to an SRE role to work in Cloud on Infra end & later infrastructure/devops 💻 2022: Started working as a cloud Developer Advocate @ AWS 💻 2023: When chatGPT launched late 2022 overtime that year upskilled and kept experimenting 💻 2024: my role at AWS officially focused on AI developer advocacy & AWS genAI services + kept upskilling in AI/ML, AI engineering 💻2025: Started to work as a Staff AI infra developer advocate at Anyscale (distributed computing). This was also a shift from corporate to startup for me and I wanted to go to the AI infra end. Currently get to work on Ray which is an open source distributing compute engine that is especially suited for AI/ML workloads (Batch Inf, Distributed Training, etc..) Relevant tags: [ AI engineer, career journey, machine learning, cloud computing, programmer, computer science, devops, MLOps, coding, setup inspiration, tech career ]
49.4K
823
70
8mo ago
lindavivah
The beauty about tech is there is no one path. 🌱 One of the biggest lessons throughout my journey has been that what felt like the hardest parts often involved “figuring out my next growth step”. The decision is sometimes harder than the execution. ⚙️ Overtime, I learned those phases of “going in circles” are totally normal & actually a lot of the work within your journey. 🔁 For example, it took me a while to figure out that I would like to go in to SRE from webdev etc. 🔁 the same thing happened when I decided it was time to go to a startup and the AI infra end. Have you experienced something similar? Lmk in comments 🤎 ⬇️ 💡 Remember, YOU are the artist of your own life 🎨 Summary of my Untraditional Journey in to Tech & Cloud & AI: 💻 2013: graduated BA in Philosophy & started working in media 💻 2015: started teaching myself how to code & a few months later attended a coding bootcamp (@flatironschool Fullstack Web Dev immersive) 💻 2016: started working as a javaScript Web Developer in media 💻 2018: Started teaching myself Cloud Computing via certs & projects, specifically AWS 💻 2020: Shifted in to an SRE role to work in Cloud on Infra end & later infrastructure/devops 💻 2022: Started working as a cloud Developer Advocate @ AWS 💻 2023: When chatGPT launched late 2022 overtime that year upskilled and kept experimenting 💻 2024: my role at AWS officially focused on AI developer advocacy & AWS genAI services + kept upskilling in AI/ML, AI engineering 💻2025: Started to work as a Staff AI infra developer advocate at Anyscale (distributed computing). This was also a shift from corporate to startup for me and I wanted to go to the AI infra end. Currently get to work on Ray which is an open source distributing compute engine that is especially suited for AI/ML workloads (Batch Inf, Distributed Training, etc..) Relevant tags: [ AI engineer, career journey, machine learning, cloud computing, programmer, computer science, devops, MLOps, coding, setup inspiration, tech career ]
Google engineer explains Agent Evaluation vs Model Evaluation in 30 seconds ⚡ 🚙 Think of getting your driver’s license. First there’s the theory test, where you’re assessed on knowledge and getting the right answers. Then there’s the road test, where you’re out in the real world actually using tools and doing things to get around. That sequence of steps out on the road has a name. It’s called the trajectory, and evaluating it is one of the things that makes agent evaluation different from model evaluation. Getting an agent from your laptop to production has stayed stubbornly fragmented, which is exactly what Agents CLI is built to fix. It works with your coding agent (Claude Code, Gemini CLI, Codex) across the whole lifecycle: It’s the programmatic backbone for the whole agent lifecycle, and it works with your coding agent (Claude Code, Gemini CLI, Codex): 🔹 Scaffold a new agent project 🔹 Build on ADK, Google’s open source agent framework 🔹 Evaluate: generate runs, grade them, optimize, repeat 🔹 Deploy and add observability 💡 Once you’ve built and deployed your agent, it can integrate right into the Gemini Enterprise Agent Platform, Google’s platform to build, scale, govern, and optimize enterprise-grade agents. That’s where it runs and scales across your org. Comment ✨AGENTS✨ & I’ll DM you the links Thank you to @pier_paolo28 for joining me on a Tech Walk and breaking this down so clearly!
42.9K
1.26K
61
3mo ago
lindavivah
Google engineer explains Agent Evaluation vs Model Evaluation in 30 seconds ⚡ 🚙 Think of getting your driver’s license. First there’s the theory test, where you’re assessed on knowledge and getting the right answers. Then there’s the road test, where you’re out in the real world actually using tools and doing things to get around. That sequence of steps out on the road has a name. It’s called the trajectory, and evaluating it is one of the things that makes agent evaluation different from model evaluation. Getting an agent from your laptop to production has stayed stubbornly fragmented, which is exactly what Agents CLI is built to fix. It works with your coding agent (Claude Code, Gemini CLI, Codex) across the whole lifecycle: It’s the programmatic backbone for the whole agent lifecycle, and it works with your coding agent (Claude Code, Gemini CLI, Codex): 🔹 Scaffold a new agent project 🔹 Build on ADK, Google’s open source agent framework 🔹 Evaluate: generate runs, grade them, optimize, repeat 🔹 Deploy and add observability 💡 Once you’ve built and deployed your agent, it can integrate right into the Gemini Enterprise Agent Platform, Google’s platform to build, scale, govern, and optimize enterprise-grade agents. That’s where it runs and scales across your org. Comment ✨AGENTS✨ & I’ll DM you the links Thank you to @pier_paolo28 for joining me on a Tech Walk and breaking this down so clearly!
⚡What are Declarative Pipelines in < 90 seconds? Let’s walk with Michael Armbrust, Distinguished Engineer at @databricksinc who led the design & development of Declarative Pipelines + created Spark SQL, Structured Streaming, Delta Lake, & much more as he breaks it down! 💡 TLDR: instead of telling the system HOW to do every step, you just tell it WHAT you want to exist. Spark figures out the how. The technology itself is open source in Apache Spark, and Databricks offers a managed version on top called Lakeflow Spark Declarative Pipelines. 💻 Want to get started with Declarative Pipelines? Comment ✨Databricks✨ & I’ll send you the resources: an intro hands-on tutorial + Michael’s recorded deep dive talk from the Data + AI Summit 2026. It was such a blast getting to chat with Michael! Did you know he was also the 9th employee at Databricks?! As always, Happy Building! 😊 #DatabricksPartner
41.5K
818
85
4mo ago
lindavivah
⚡What are Declarative Pipelines in < 90 seconds? Let’s walk with Michael Armbrust, Distinguished Engineer at @databricksinc who led the design & development of Declarative Pipelines + created Spark SQL, Structured Streaming, Delta Lake, & much more as he breaks it down! 💡 TLDR: instead of telling the system HOW to do every step, you just tell it WHAT you want to exist. Spark figures out the how. The technology itself is open source in Apache Spark, and Databricks offers a managed version on top called Lakeflow Spark Declarative Pipelines. 💻 Want to get started with Declarative Pipelines? Comment ✨Databricks✨ & I’ll send you the resources: an intro hands-on tutorial + Michael’s recorded deep dive talk from the Data + AI Summit 2026. It was such a blast getting to chat with Michael! Did you know he was also the 9th employee at Databricks?! As always, Happy Building! 😊 #DatabricksPartner

Linda Y (@lindavivah) Instagram Stats & Analytics

Linda Y (@lindavivah) has 116K Instagram followers with a 1.41% engagement rate over the past 12 months. Across 148 posts, Linda Y received 142K total likes and 10.6M impressions, averaging 962 likes per post. This page tracks Linda Y's performance metrics, top content, and engagement trends — updated daily. To benchmark @lindavivah against other accounts, use Socialpruf's Instagram competitor analysis tool.

Linda Y (@lindavivah) Instagram Analytics FAQ

How many Instagram followers does Linda Y have?+
Linda Y (@lindavivah) has 116K Instagram followers as of October 2026.
What is Linda Y's Instagram engagement rate?+
Linda Y's Instagram engagement rate is 1.41% over the last 12 months, based on 148 posts.
How many likes does Linda Y get on Instagram?+
Linda Y received 142K total likes across 148 posts in the last 12 months, averaging 962 likes per post.
How many Instagram impressions does Linda Y get?+
Linda Y's Instagram content generated 10.6M total impressions over the last 12 months.
Can I compare Linda Y with other Instagram accounts?+
Yes. Add @lindavivah to a competitor set with Socialpruf's Instagram competitor analysis tool and benchmark it side by side with any other public accounts on views, engagement rate and follower growth, with history back to each account's first post.