Building AI products & intelligent workflows
Coimbatore
Joined July 2026
- Tweets51
- Following9
- Followers5
- Likes84
A Gemini warning taught me something important about LLMs:
LLMs are stateless. Your code keeps the memory.
AFC lets the SDK handle the tool-calling loop, while Chat keeps the conversation history for follow-ups.
Tool descriptions are prompts too.
#AIEngineering #LLM #Gemini
Valid JSON ≠ usable JSON.
An LLM can return perfectly valid JSON and still break your application.
If your code consumes the output, don’t just ask for JSON. Define the schema and enforce the types.
Prompts describe the format. Schemas enforce it.
#LLM #AIEngineering #GenAI
Talking about LLM temperature 🌡️
A small setting that affects how predictable or varied a model’s responses can be.
Lower → more predictable
Higher → more varied
The exact range depends on the model/provider.
Happy to share my latest achievement: Claude Certified Architect - Foundations certified by Anthropic! 🎓 Excited to put these AI architecture principles into practice.
#BuildWithAI #Claude #AI #TechMilestone
Took a step back from coding today and went back to the basics of AI.
Revisited ML, LLMs, Generative AI, RAG, Agentic AI, and the tools behind them.
Sometimes, understanding the “why” behind what we build is just as important as making it work.
docs.google.com/document/d/1…
Never thought the route could make this much difference.
Something worth knowing before your next swap. ⚡️
I didn't think route mattered until I checked the same swap on two platforms and got two different numbers
@swapdotio checks 5+ aggregators for you and picks the best one automatically
so you're not the one comparing tabs every time
Real swap, real numbers, in the video
If you're swapping on @solana
this is worth 30 seconds of your time
@superteam
Baddies aren't apolitical.
Documenting voices and perspectives from Freedom Park, Bengaluru, on the NEET debate. Listen to what people have to say and decide for yourself.
#NEET #Bengaluru #FreedomPark
Dharani Mahendiran retweeted
SuperteamIndia🇮🇳
@SuperteamIN @superteam @usualsai
Communities don't grow because of hype.
They grow because people keep showing up.
Every hackathon. Every meetup. Every late-night shipping session.
This video is my take on the people behind @solana builders, creators, and storytellers who make the ecosystem what it is.
With @usualsai , every builder has the chance to share their journey, showcase their work, and leave their mark on the community through Global Solana Faces. Because every contribution—big or small—deserves to be seen.
Dont just build ,be a part of the story
Hi everyone! 👋 I'm new here, and I wanted to share one of my favorite projects.
A few months ago, I built something that means a lot to me.
Not because it was just another website, but because it was for my mom's cloud kitchen.
From the initial idea to the design, development, and launch, I built the entire website from scratch.
Seeing something you built help someone you care about is a different kind of satisfaction.
I'd love for you to take a look, and I'd really appreciate any feedback.
🌐 lachoparatha.in
If this line of research holds up, it could reshape how we think about reasoning, memory, and transparency in language models. Excited to see where it leads.
This is a really interesting perspective. Feels like AI is moving from being just a tool to becoming more of a teammate. Excited to see where this goes.
Dharani Mahendiran retweeted
Claude Fable 5 will be available again globally tomorrow.
After a series of productive conversations with the US government, we're redeploying the model with a new set of classifiers to target and block more cybersecurity tasks. In the near term, some routine tasks like coding and debugging will fall back to Opus 4.8. We’ll continue to refine these classifiers over the coming weeks to reduce false positives and better distinguish genuine misuse from legitimate requests.
We’ve also begun drafting a consensus framework—with Amazon, Microsoft, Google, and other Glasswing partners—for assessing the severity of AI jailbreaks and how AI developers should respond to them. We invite other industry partners and model providers to join us in this effort.
Finally, we’re scaling up our collaboration with the US government on model testing and safeguards. This will include pre-release access to models and safeguards for evaluation, information sharing on jailbreaks and misuse, and dedicated resources for joint research.
Thank you to our users for your patience, and to our partners across the government, industry, and the research community who worked alongside us to make Fable 5 available again.
Read our full blog: anthropic.com/news/redeployi…
Dharani Mahendiran retweeted
We’re shipping two major updates to streamline your creative workflow, allowing you to generate high-speed images with one model and then instantly animate them with the other—all at a fraction of the cost 🍌⚡️
1️⃣ Introducing Nano Banana 2 Lite: Our fastest and most cost-efficient Gemini Image model yet delivers text-to-image outputs in under 4 seconds. Now available via the Gemini API and Google AI Studio, and rolling out soon across @NotebookLM, @FlowbyGoogle, @geminiapp, @stitchbygoogle, Google Search and @GooglePhotos.
2️⃣ Gemini Omni Flash in Public Preview: Our natively multimodal model for cost-efficient video generation and conversational editing. Now available via the Gemini API, @googleaistudio, and Gemini Enterprise Agent Platform so you can integrate the model into your workflow.
While exciting on their own, the real magic happens when you build using these models together.
Watch how our interior design demo integrates Nano Banana 2 Lite and Omni to instantly reimagine any space. Upload a photo, swipe through tailored design concepts, and see Omni bring the details to life in cinematic motion.
Try out the demo app in AI Studio: goo.gle/443xPqw