Hello World: A New Beginning

Welcome to my new website. I’m Karthik Bhaskar, a Senior Applied Scientist at Vijil.ai, where I work on trust infrastructure for AI agents.

Why a new site?

I’ve been working in AI/ML for several years - from building recommendation systems at the University of Toronto, to shipping enterprise AI at CIBC, to co-founding startups. Now, as I focus on trustworthy AI research, I wanted a space to share thoughts more openly.

What to expect

I plan to write about:

• Adversarial ML - attacking and defending language models
• AI Safety - making AI systems reliable in production
• LLM Security - prompt injection, jailbreaks, and defenses
• Applied Research - bridging the gap between papers and products

A code example

Here’s a simple Python snippet to demonstrate syntax highlighting:

def trust_score(model, test_suite):
    """Evaluate model trustworthiness across a test suite."""
    results = []
    for test in test_suite:
        response = model.generate(test.prompt)
        score = evaluate_safety(response, test.expected)
        results.append(score)
    return sum(results) / len(results)

Stay tuned for more posts. You can find me on GitHub, LinkedIn, or X.