Open to opportunities

Grishav Paudel

@grishavpaudel

Message

Data scientist and agentic AI engineer building LLM-powered systems that deliver business value.

India

Message

What I'm looking for

I’m looking for an early-career role to design LLM/agentic AI systems and data pipelines, collaborate with product teams, and prototype-to-deploy models that improve business decision-making and learning outcomes.

I’m a Data Scientist and agentic AI engineer who builds end-to-end AI solutions—from data pipelines to LLM integrations—focused on real decision-making impact. At PHOENIXFLY-A-WAY PVT. LTD., I designed AI-driven data pipelines and analytical models, and I helped shape AI architecture choices like model selection and deployment strategies while integrating LLM-based features into product workflows.

Recently, I scaled prompt-based content and adaptive learning experiences as a Prompt/Context Engineer, creating an automated prompt pipeline that cut manual CUET content creation by ~60% and deploying a full-stack exam platform for 500+ students with adaptive question selection and real-time scoring. I’ve also delivered forecasting and applied ML work (e.g., ~88% traffic prediction accuracy at UBER), and I build production-ready projects like a voice + chat immigration agent with safety guardrails, a multimodal RAG system for educational Q&A with cited responses (~87% retrieval accuracy), and a real-time fraud detection pipeline using streaming ML with <100ms latency.

Experience

Work history, roles, and key accomplishments

Current

Data Science & AI Intern

Current

Phoenixfly-A-Way Pvt. Ltd.

Feb 2026 - Present (4 months)

Designed and developed AI-driven data pipelines and analytical models to support business intelligence and core operations decision-making. Helped shape AI architecture decisions by selecting models, defining deployment strategies, and integrating LLM-based features into existing product workflows.

Data Pipelines AI Integration AI Architecture AI Model Deployment Prompt Engineering Business Intelligence Analytics System Integration

Prompt/Context Engineer

Unimonks

Oct 2025 - Jan 2026 (3 months)

Built a structured prompt pipeline using few-shot prompting to auto-generate illustrated CUET study content, reducing manual creation time by ~60%. Developed and deployed a full-stack AI exam platform with adaptive question selection and real-time scoring, using system prompts and evaluation rubrics to reduce hallucinations across 10+ subject domains.

Prompt Engineering System Prompts Real Time Scoring

Data Science Intern

Uber

Jul 2025 - Aug 2025 (1 month)

Developed a traffic volume forecasting model using XGBoost and time-series feature engineering on junction-level history, achieving ~88% prediction accuracy (RMSE ~120 vehicles/hour). Owned end-to-end pipeline work including cleaning, feature engineering, tuning, and visualization across 4 major road junctions.

Time Series XGBoost Feature Engineering Data Cleaning & Model Tuning RMSE Evaluation Data Visualization Jupyter