A comprehensive guide to Palantir Foundry's Ontology strategy. / 世界最強のデータプラットフォーム「パランティア」の中核概念である『オントロジー』の戦略と実装を解き明かすOSS書籍プロジェクト。

Practice #ontology engineering with Python (and other languages).

compile docs and github reference code, custom validators, maybe convex to osdk

From MVP to business value: explainable AML AI ready for real-world deployment.

Long-form posts on AI, ML, Data Engineering and technology - in collaboration with my team

End-to-end CRM data model inspired by Palantir Foundry. Includes raw datasets, cleaning pipelines, ontology modeling, lineage, and dashboard.

Backend for the AI Interview Mentor platform. FastAPI server providing REST APIs for user authentication, resume upload, interview sessions, and analytics. All data is securely stored and managed in Palantir Foundry via Ontology SDK. Acts as a secure proxy for Palantir operations and keeps secrets off the frontend.

DSPY for model agnostic optimization. Dockerized deployment of FastAPI classification routes to railway, cf workers, and Palantir Foundry. Pharma/Medtech complaints vs. AE example.

Tenant churn prediction for single-family rental management — XGBoost ML pipeline with Palantir Foundry AIP workflows, Snowflake analytics, and real-time retention dashboard

mvp of streaming pipelines by emulating a trimmed version of the palantir txn and dataset APIs. Includes an AI search / enrichment flow using gemini + google search. written in go and with no pre-caching optimizations seeing 10x faster start up for full local testing.

DeepSigma turns AI decisions into auditable, recoverable infrastructure. It seals each decision episode end-to-end (inputs, actions, outcomes), detects drift, enforces governance gates, and drives patch loops with measurable confidence. Built in Python with MCP, LangChain, and OpenTelemetry.

Analysis of post-2000 US foreign aid on democratic outcomes