Aaseya

Happy to share that I’ve completed the Claude 101 certification course by Anthropic Claude . Excited to keep exploring AI, prompt engineering, and real-world applications of LLMs. 🚀 #AI #ClaudeAI #Anthropic #Learning #Certification

I finally know what to ask AI, how to ask it… and when to double‑check it :-) #AIFluency

Happy to share that I’ve joined IQVIA as SDE 2 . Excited for this new chapter—looking forward to learning, building, and creating meaningful impact along the way. 🚀

Allbirds — the sustainable sneaker brand that once defined Silicon Valley's tech-casual uniform — is now rebranding as NewBird AI. Just days before shutting down operations, the company announced a pivot to AI computing infrastructure. They've secured $50M in financing, plan to acquire high-performance GPUs, and will offer GPU-as-a-Service to AI developers and enterprises. The market reacted instantly. Allbirds' stock surged ~600-700% overnight — from under $3 to over $16/share. Is this a genius pivot into a booming AI infrastructure market, or a desperate AI-washing play by a dying brand? On one hand: GPU demand is exploding globally Lead times are long, and hyperscalers can't serve everyone There's genuine need for dedicated GPU-as-a-Service On the other: Allbirds had ZERO history in AI, semiconductors, or cloud infrastructure They sold their entire footwear business for just $39M — a fraction of their $4B peak valuation Rebranding from a $4B sneaker brand to a $50M AI startup... is that transformation, or a last resort? The AI infrastructure gold rush is real. But so is the "AI-washing" epidemic — companies slapping "AI" on their name to chase stock momentum. As AI leaders, what's your take? Is Allbirds' move a bold reinvention worth watching — or the most transparent AI pivot of the decade? I'd love to hear your thoughts. #AI #ArtificialIntelligence #TechPivot #GPUaaS #StartupStrategy #BusinessStrategy #Allbirds #NewBirdAI

Excited to complete the Anthropic Claude 101 Certification and deepen my understanding of modern AI

Most enterprises are burning through LLM token budgets without realising the real culprit is not the model — it is poor memory architecture. Fine-tuning a Large Language Model without a disciplined memory ops strategy means every redundant context window, bloated prompt chain, and unmanaged KV-cache hit is silently inflating your inference cost. Memory optimisation is not a DevOps afterthought — it is the financial lever that determines whether your AI investment scales or bleeds. For Agentic AI, this becomes existential. Autonomous agents operate across multi-step reasoning loops and long-horizon tasks. Without intelligent memory tiering — working memory, episodic recall, semantic retrieval — your agents are re-processing what they already "know," wasting tokens on every cycle. Enterprises that treat memory ops as a first-class citizen in their LLMOps stack will extract far more value per dollar: leaner fine-tuning runs and agents that compound intelligence rather than reset it. The question is not whether you can afford to invest in memory optimisation. It is whether you can afford not to. How is your organisation approaching memory rationalisation in your LLM and Agentic AI stack? Drop your thoughts, challenges, or unconventional ideas below — let us build this knowledge together. #LLMOps #AgenticAI #MemoryOptimisation #EnterpriseAI #GenerativeAI #AIArchitecture #LargeLanguageModels #TokenOptimisation #AIStrategy #FutureOfWork