umesh

[amit @amitisinvesting]

$IBM

Woah.

IBM put out preliminary Q2 results...and they are not good.

Revenue was $17.2B vs $17.86B expected, up 1% YoY. Consulting revenue was flat and infrastructure revenue was down 7% YoY.

All the software names are down in sympathy to IBM basically saying that software spend went to semiconductor spend this quarter.

CEO Arvind Krishna: “In the last few weeks of June, we saw clients shift their quarterly capex spend toward servers, storage, and memory purchases to secure supply-constrained infrastructure ahead of expected price increases.” He also said clients were distracted by “rapidly-evolving, industry-wide cybersecurity concerns,” and that IBM “did not anticipate the magnitude of the capex reprioritization.”

Company just lose $65B in market cap.

I don't think all the software names should be down because of IBM's specific news but today could be a day where the fear of this happening to other names affects all of them.

$IBM -24%

[umesh @umianta]

Microsoft last week dropped a new markup language for LLM prompts to solve common prompt engineering challenges. 🚀

I built a hands-on lab + walkthrough video to explore it 👇

💻 Code: https://t.co/L1vbbRTuI0
🎥 Video: https://t.co/9IcLqvNFRq

#AI #LLM #PromptEngineering

[umesh @umianta]

Build, Deploy, and Test an MCP (Model Context Protocol) Server | MCP In... https://t.co/ZHRydbaLsC via @YouTube
#MCP #llm

[umesh @umianta]

Control LLM model using Guidance-AI
https://t.co/ijIEdKItkM

[umesh @umianta]

Phi-4-Mini: Compact Multimodal Language Models via Mixture-of-LoRAs

https://t.co/vdR6MNvi01

[umesh @umianta]

Deepseek research paper
https://t.co/wEv4DCfgAn
#DeepSeek #Ai #LLMs #genai

[umesh @umianta]

How LLM Models Think When Solving Math Problems https://t.co/XYRTIQMoAQ via @YouTube

[umesh @umianta]

Set Up & Run DeepSeek-R1Model Locally: Step-by-Step Guide Using Open Web... https://t.co/Ha4AReYozu via @YouTube

[umesh @umianta]

OpenAI's Real-Time API​ - Unlock Low-Latency and Multi-Modal Conversatio... https://t.co/3pN0y5XdXj via @YouTube

[umesh @umianta]

Exploring Multi-Modal RAG In Python - Step by Step Demo https://t.co/kZbAtfLybO via @YouTube

[umesh @umianta]

https://t.co/xf0ghBPcLt

[umesh @umianta]

Conversational Bot and AI Phone Agent Phone Integration https://t.co/0PuYgGpYOW via @YouTube

[umesh @umianta]

Capture telemetry data from Copilot Studio Chatbots using Azure Applicat... https://t.co/7OrsgqiiNh via @YouTube

[umesh @umianta]

Watch my video below to delve deeper into Copilot Studio plugins.
In this video, you'll discover:
✅ The concept of Copilot Studio plugins
✅ Constructing AI Builder prompt plugins
✅ Integrating and leveraging them within the Copilot Studio Chatbot
https://t.co/McaoDKlaf5

[umesh @umianta]

I experimented with Small LLMs and Retrieval-Augmented Generation (RAG) chain that operates locally on my machine. In the design of this Chatbot design, you have the flexibility to input any website URL along with questions. https://t.co/sxcN1o4Er0

[umesh @umianta]

I created a comprehensive design featuring a Chatbot & AI Phone calling hyper-AI capabilities. This design involves an AI-powered chatbot that interacts with the AI phone calling platform to initiate a call. #genai #openai #chatbot #ai #micorsoft #powerautomate

[umesh @umianta]

📷I have built an RAG Arch-based Chatbot by leveraging Pinecone Vector DB, Open AI embedding & summarization models, MS Copilot studio & Power flow.
To explore the code design, you can visit my GitHub repository here -https://t.co/d9JVdCoRzv
Here's the architectural diagram -

[umesh @umianta]

@"Conversational AI Chatbot(PVA)- New Advanced AI Features "https://t.co/naddKsm7tW on @LinkedIn

[umesh @umianta]

@"Azure Cloud Monitoring Solutions"https://t.co/bao70soXZL on @LinkedIn

[umesh @umianta]

@"Access GPT Models using Azure OpenAI"https://t.co/hGZp6v6YjF on @LinkedIn @azure