At Wamolabs, I architect and deploy LLM-powered and agentic AI systems for enterprise workflows. I also built Lingopal, a real-time speech-to-speech translation platform using streaming pipelines with Whisper and ElevenLabs.
For Ascendia, I designed a multi-agent recruitment platform with CrewAI, FastAPI, and vector search. Its workflows reduced manual screening effort by 50%.
Previously, at Devsinc, I developed LLM-powered APIs and agentic workflows, and fine-tuned domain-specific LLMs using QLoRA. At Developers Core, I trained and deployed deep learning models with PyTorch and TensorFlow and built end-to-end ML pipelines.
My project work also includes an enterprise knowledge-discovery agent system for Recursion and an AI-powered SEO automation platform. I work across machine learning, generative AI, and production deployment.
