Just launched my digital store!.My first product is live: “30 Days of Daily Devotionals for Faith-Driven Leaders.”. It’s free to download — or you can pay what you feel it’s worth. I’d love your feedback as I grow this store one resource at a time.

RT @jjude: Responsible AI: In theory vs in practice. I spoke at the NASSCOM CXO Breakfast session in Chandigarh on how we’re actually appl….

RT @sandeep_PT: From scratch, a ravaged nation was built by the toil of crores of Indians from the 1950s. Brick by brick. Project by projec….

RT @AditiRajasthan: The most valuable math you can learn.is how to calculate the future cost of your current decisions.

RT @malpani: You can rate potholes at

You can read the edited transcript of the talk here: 11/11.

Responsible AI is not a checkbox. It’s culture. And like all culture shifts, it needs leadership, repetition, and cross-functional effort. 10/11.

Responsible AI ≠ perfect AI.It means mature systems, clear policies, and real-world awareness. You won’t fix foundational issues. But you can build responsibly at the application layer. 9/11.

Apply RMF again:. 🗺️ Map: what can it say? where’s the data from?.📊 Measure: who’s asking what? token usage? risky replies?.🔧 Manage: alerts, logs, response controls.📋 Govern: review cycles, RACI, incident playbooks. 8/11.

External AI apps (chatbots, sales tools)? Different ballgame. Now you're facing customers. So governance must go deeper. Example: a chatbot should be clear if it's answering from a KB vs internet. 7/11.

Internal AI usage (employee tools, SDLC copilots, etc.):.✅ Usage policy.✅ Tool whitelist.✅ Training & repetition.✅ Feedback loops between engineers and CXOs. Keep it lightweight, but intentional. 6/11.

But that doesn’t mean we throw up our hands. We apply the NIST AI Risk Management Framework (RMF):. • Map.• Measure.• Manage.• Govern. It works for both internal and external use cases. 5/11.

Most of us are building on shaky foundations. OpenAI, Claude, Perplexity—they all use datasets we don’t fully control or understand. You can’t guarantee fairness or privacy when you don’t know what went into the model. 4/11.

It’s like asking an Indian exec: “Should we bribe?”.Everyone says no. Then ask: “Have you bribed?”.Most say yes. Welcome to the real-world tension between principle and practice. 3/11.

In theory, Responsible AI sounds great. • Fairness. • Explainability. • Transparency. But when you’re racing to ship GenAI apps, ideals often crash into growth goals. 2/11.

Responsible AI: In theory vs in practice. I spoke at the NASSCOM CXO Breakfast session in Chandigarh on how we’re actually applying Responsible AI while building enterprise apps. 👇. 1/11

> The problem of good code is to solve a specific set of problems efficiently. AI hasn't perfected this yet. It solves problems, sure, but not always efficiently. Very good take on AI code generation. This is what I have seen in my experience too.

Things take longer to happen than you think they will, and then they happen faster than you thought they could. — Rudiger Dornbusch.

RT @NoahRyanCo: Build your life around your strengths and build systems around your weaknesses. Write this down.

RT @SahilBloom: Major career cheat code: Be the person who can figure it out. You’ll be handed tasks you have no idea how to complete. Ther….

RT @Soumya_Ray: @svembu A school decides it is going to focus on the top of the class to show best results, rather than try and get the wea….