
Priyanka Vergadia
@pvergadia • 48,494 subscribers
15 Years building AI and Cloud @microsoft @google @intel | #1 Visual Storyteller in Tech | TED Speaker | DM for Collabs | Opinions = mine
Shorts
Videos

BREAKING: Microsoft just launched its own AI model family and the coding model inside GitHub Copilot. MAI-Code-1-Flash. Shipped to every Copilot tier. I just tested it to build an app. Check out the video 👇 85.8% benchmark score. 256K context. 60% token reduction on hard tasks. Here are the other models. → MAI-Code-1-Flash (coding, free) → MAI-Thinking-1 (reasoning, 35B params) → MAI-Image-2.5 (image generation) → MAI-Transcribe-1.5 (speech, 43 languages)
Priyanka Vergadia37,018 views • 1 month ago

This girl literally explains the AI Bubble in 12 mins! (Who is this girl?!😀)
Priyanka Vergadia76,990 views • 6 months ago

JP Morgan revealed the exact architecture behind "Ask David" their investment research AI agent. Supervisor. Subagents. Judge. Human. Here's the whole stack. → 1 supervisor agent routes and orchestrates → Specialized subagents handle retrieval, structured data, and analytics separately → LLM-as-judge reflection runs before any answer leaves the system → Human-in-the-loop patches the last mile accuracy gap AI can't solve alone Every serious agent team is converging on this pattern. JP Morgan just confirmed it works at enterprise scale. Build accordingly. worth watching 👇
Priyanka Vergadia17,075 views • 2 months ago

In my recent interactions with CTOs and Eng managers, this is coming up a lot: The SDLC is changing shape. We are moving from the "Middle Heavy: Coding" era to the Hourglass. Let me explain! Here is the breakdown of the Agentic SDLC: 🔻 THE TOP: INTENT (Heavy) The "Source Code" is no longer the syntax; it’s the specs, the docs, and the customer signals. Why: AI executes literally. Vague intent = Broken product. The Job: Context Engineering. 🔺 THE MIDDLE: BUILD (Collapsed) The actual coding phase is becoming invisible infrastructure. Why: AI Agents handle the syntax, the grind, and the boilerplate. The Job: Supervision, not translation. 🔺 THE BOTTOM: VERIFICATION (Heavy) This is the new constraint. Traditional Peer Reviews cannot keep up with AI speed. Why: You can't manually review 1,000 lines of generated code in 10 minutes. The Job: Building automated guardrails and safety checks. The Bottom Line: Treat your Project Management workspace like your compiler. If the input (intent) is buggy, the output (software) will be broken—faster than ever before. #EngineeringManagement #CTO #AI #SoftwareDevelopment #Agile #SDLC
Priyanka Vergadia18,085 views • 6 months ago

Here is a quick look at the newly announced #DuetAI for Google Cloud which is an always-on AI collaborator that uses generative AI to: 👍 provide contextual code completion 👍 generate entire functions in real-time 👍 provide suggestions tuned to your code base 👍 assist you with code reviews and inspections 👉 learn more 👉 Wanna try? Sign up #GenAI #GoogleIO
Priyanka Vergadia12,808 views • 3 years ago
No more content to load