1:1 AI/LLM Engineering & Strategy Consultation

Book a 60-minute 1:1 with me if you're building with LLMs, AI agents, or ML infrastructure and want direct, technical guidance — not generic advice.

I'm currently Founding CTO at PromptGuard. Before that I built a 30B-parameter LLM from scratch at Multiverse Computing, shipped voice search serving 200M+ daily users at Flipkart, and co-authored CompactifAI (ESANN 2025), a compression method that cut LLM size 70% with under 3% accuracy loss.

What we can cover:

• LLM training and fine-tuning (FSDP, DeepSpeed, LoRA, DPO)
• Inference optimization and serving (vLLM, TensorRT, quantization)
• Model compression and cost reduction
• Agentic system architecture and safety
• Technical strategy for early-stage AI startups