Headline

ML Engineer | ex-Snap| ex-Meta

About

I’m a Machine Learning Engineer passionate about decoding the brain using AI. With a background in neural time-series (EEG), multimodal data fusion, and deep learning, I work on building interpretable models to understand meditative brain states, cognitive processing, and more. Currently working on: 🧠 Real-time EEG classification of meditative vs non-meditative states 🧬 Applying foundation models to biological data 🤖 Aligning neuroscience with large-scale ML (ViTs, Transformers, CEBRA) I’m particularly drawn to projects that push the frontiers of neuro-AI by scaling transformer models for neural code modeling. I love working across ML pipelines — from raw neurophysiological data to scalable training systems — and seek to build tools that bring us closer to understanding intelligence.

Location

Bengaluru

AI stack

claude code

Work experience

Linguist II at Meta, Sep 2025 - Jul 2026 · 11 mos, Burlingame, California, United States, ● Served as Language Lead for Ray-Ban Meta Glasses, owning the end-to-end semantic and linguistic quality strategy for multimodal, voice-driven AI experiences spanning audio, vision, and text. ● Designed and led a multimodal LLM evaluation architecture, integrating offline metrics, structured human review, and production telemetry to detect hallucinations, semantic drift, and intent loss. ● Defined shared error taxonomies and gold evaluation datasets adopted across research and product teams, establishing a consistent quality contract for multilingual AI behavior. ● Set technical direction for human-in-the-loop workflows, explicitly defining automation boundaries and confidence thresholds to balance latency, cost, and creative accuracy. ● Partnered with ML research and platform teams to translate evaluation findings into architectural improvements, influencing prompt design, data curation, and safety guardrails.

Principal Engineer and Co Founder at MoodMe, Aug 2025 - Feb 2026 · 7 mos, Belgium, - Lead hands-on development of ML and computer vision product features across the full SDLC, with focus on mobile and edge SDKs, face and body analytics, and augmented reality experiences. - Drive technical execution for MoodMe products by designing maintainable software, improving code quality, and applying DevOps best practices across development workflows. - Collaborate with product, design, and executive stakeholders to translate product concepts into launch-ready AI features and user-facing capabilities. - Mentor junior engineers, provide technical guidance, and support team productivity through code reviews, onboarding, and structured execution. - Research and apply emerging ML, computer vision, and software engineering practices to improve product performance, user experience, and technical scalability.

Machine Learning Engineer at Snap Inc., Dec 2023 - May 2025 · 1 yr 6 mos, Designed and shipped multimodal ML systems for AR experiences that fused visual, textual, and contextual signals into real-time semantic understanding and summarization. Architected LLM-powered agentic workflows with retrieval, tool use, and structured outputs, reducing end-to-end inference latency by 45% in production settings. Built shared RAG and orchestration foundations that standardized retrieval, evaluation, and observability patterns across teams, reducing duplicated development effort. Implemented reliability guardrails and evaluation pipelines using NDCG, BLEU, and domain-specific quality metrics to improve output accuracy and reduce hallucination risk. Influenced deployment architecture across quantization, distributed inference, and throughput optimization, improving performance by 25% while preserving user experience quality.

Machine Learning Engineer at Clinet it solution, Aug 2022 - Aug 2023 · 1 yr 1 mo

Machine Learning Engineer at _VOIS, Jul 2022 - Sep 2022 · 3 mos, Karnataka, India, • Orchestrated the development of a 97% accurate model using Azure Machine Learning, from data ingestion with Azure Data Factory to refinement using Azure Databricks. Ensured robust validation in collaboration with cross-functional teams. • Elevated software QA processes by 98% through strategic implementation of Azure DevOps and Azure Data Lake, optimizing workflow efficiency and data integrity.

Machine Learning Engineer Intern at Clinet it solution, Aug 2020 - Aug 2022 · 2 yrs 1 mo, Bengaluru, Karnataka, India, ● Built and deployed a personalized financial product recommendation engine using unsupervised learning (K-Means, DBSCAN) and collaborative filtering (SVD, ALS), which boosted client engagement by 25%. ● Performed detailed feature engineering on over 50M+ transaction logs and user activity events, creating behavior-based customer embeddings to power clustering and similarity scores, enhancing model interpretability and recommendation precision. ● Designed a hybrid recommendation pipeline that combined content-based profiling with user-item collaborative signals, enabling cold-start problem mitigation and expanding recommendation coverage by 35% across new user segments. ● Containerized the system as a scalable ML microservice using FastAPI, Docker, and Azure App Services, reducing inference latency by 45% and enabling seamless integration with Twinkle's client-facing mobile app. ● Established a robust evaluation framework using A/B testing, precision@k, recall, and NDCG, and implemented model retraining every 7 days with Airflow + Azure Blob Storage to keep recommendations aligned with user behavior drift.

Education

Master of Science - MS, Data science, Stevens Institute of Technology

Bachelors-Technology, Computer and Information Technology, Reva University