Open to opportunities

Pranjal Raghava

@pranjalraghava

Message

Applied AI engineer designing LLM reasoning pipelines and multimodal human-understanding models.

India

Message

What I'm looking for

I seek roles building production-grade AI systems (LLM/RAG, multimodal inference) with strong MLOps, interpretability, and impact-driven teams.

I am an Applied AI engineer specializing in designing LLM reasoning pipelines and multimodal human-understanding models, with hands-on experience building production AI systems across behavioral speech intelligence, organizational influence modeling, and multimodal health inference. I have designed and deployed RAG/Graph-RAG architectures, agentic systems, and robust MLOps pipelines using tools like LangChain, LangGraph, vector databases, Docker, Airflow, and Kubernetes.

My work includes architecting CLAREX for speaker diarization and psychometric scoring, developing temporal graph inference for organizational influence across 1K+ networks, and engineering non-contact physiological signal models with state-of-the-art accuracy. I focus on interpretable, reliable AI—combining embedding systems, hallucination detection, automated retraining, and evaluation—to deliver measurable impact in production environments.

Experience

Work history, roles, and key accomplishments

Current

Data Scientist

Current

ValueMatrix AI

Sep 2025 - Present (5 months)

Designed temporal graph inference models to reconstruct organizational influence across 1K+ networks and built multimodal behavioral intelligence (CLAREX) achieving 95% speaker diarization accuracy; developed psychometric trait inference and a CatBoost job classification engine that improved accuracy by 45%.

PyTorch CatBoost NLP Embeddings Speaker Diarization Behavioral Modeling

Data Scientist

WellWiz Innovation

Feb 2025 - Jun 2025 (4 months)

Built multimodal human-state inference and LangGraph decision pipelines for health AI, developed rPPG spatiotemporal models with 1.5 bpm MAE, and architected an MLOps platform with automated retraining and a LangChain RAG chatbot with hallucination detection.

LangGraph Mediapipe Lang Chain Docker Airflow PostgreSQL

Data Scientist

Analytics Vidhya

Oct 2022 - Sep 2023 (11 months)

Developed and deployed an XGBoost malware detection system (85% recall) with Azure Docker deployment and MLflow/Grafana monitoring; built a Pinecone-backed financial RAG chatbot for querying 50K+ trades using sentence-transformers.

XGBoost Azure Docker MLFlow Pinecone Sentence Transformers Vector Search CI CD

Deputy Engineer

Bharat Electronics Limited

Feb 2021 - Jul 2022 (1 year 5 months)

Built production underwater surveillance systems fusing sonar arrays and deep learning to achieve 95% accuracy at 50ms latency and developed unsupervised sonar anomaly detection with 92% recall for early threat alerting.

Signal Processing Sonar Systems Deep Learning Real Time Systems

Data Scientist Intern

HP Inc

Jan 2020 - Jul 2020 (6 months)

Developed a federated LSTM anomaly forecasting pipeline for 3D-printing that reduced defects by 35% and engineered a real-time ETL pipeline using RabbitMQ and Celery to integrate anomaly prediction models.

Federated Learning ETL RabbitMQ Celery Anomaly Detection Time Series