"Any sufficiently advanced technology is indistinguishable from magic."
AI/ML Engineer · Data Architect · Cloud Wizard. Conjuring intelligent systems from raw data, one spell at a time.
I am Kshitish Neupane, a practitioner of the dark arts of Artificial Intelligence, Machine Learning, Data Engineering, and Cloud Architecture — disciplines considered arcane by Muggles, but second nature to those who dare to study them.
From Nepal, I build intelligent pipelines that turn raw data into prophecy, train models that predict futures, and deploy them to clouds that span continents.
Training models that see patterns no human can. From classical algorithms to deep neural networks.
Conjuring pipelines that move mountains of data with grace. ETL, streaming, warehousing.
Raising fortresses in the sky — scalable, resilient, cost-efficient. Three clouds, one wand.
The primary wand material. Python for intelligence, SQL for querying the ancient scrolls.
Keeping the magic alive in production. Model monitoring, CI/CD for ML, LLM deployment at scale.
Translating prophecies into charts mortals can comprehend. Dashboards that tell stories.
An end-to-end ML pipeline that ingests raw telemetry data, trains prediction models, and serves forecasts via REST API. Built on AWS SageMaker with Airflow orchestration, processing 5M+ records daily.
View Spellwork →Real-time streaming data platform using Kafka + Spark. Ingests IoT events, transforms into Snowflake with automated dbt models.
View Spellwork →Multi-cloud IaC framework using Terraform. Identical environments across AWS, GCP and Azure with one command.
View Spellwork →RAG-based LLM assistant fine-tuned on domain knowledge. Deployed on GCP with semantic search over 100K+ documents.
View Spellwork →Leading ML model development and data pipeline architecture. Designing scalable ETL workflows and deploying models to cloud infrastructure.
Built and maintained data pipelines processing millions of records. Migrated on-premise infrastructure to AWS. Reduced pipeline runtime by 60%.
First steps into the wizarding world of ML. Trained classification models, built preprocessing scripts, deployed sentiment analysis tool.
Foundation in algorithms, networks and software systems. Graduated with honours.
Whether you want to build something extraordinary, need a data wizard on your team, or just want to talk about Leclerc's latest qualifying lap — the owl finds me.