tb
think big. techies
webinar series
free · live · zoom

Palantir Ontology,
decoded.

a friendly 45-minute walk-through ✨
🗓 whenSat · Aug 15
🇺🇸 edt11:30 AM
🇮🇳 ist9:00 PM
Chaitanya Vullaganti — data & ontology engineer
Let me show you the ontology—it's easier than the docs make it sound 🌱
👤 Customer object type
📦 Order object type
🏷️ Product object type
Action action type
♥ 3
↑ that's the semantic layer
🎟 grab your seat

Register free.

30 seconds — Zoom link lands in your inbox ✉️

✓ free ✓ zoom invite instantly ✓ recording included
Sat · Aug 15 · 11:30 AM EDT · 9:00 PM IST
Zoom join details emailed on registration.
--
days
--
hrs
--
min
--
sec
more about the ontology below
🖼 the story

What's a Palantir ontology, anyway?

in three cartoon panels — no jargon ✨
1
? ?
Just raw data.

Tables, rows, columns. Everything's technically here but no one knows what any of it means in the business.

2
✨poof✨
Add an ontology.

Wrap the data in Object Types, Link Types, and Action Types. Suddenly it speaks your business language.

3
= happy graph! 🌱
Meaning.

Named entities with real relationships. Every app in Foundry reads from one shared, governed truth.

🧱 the three ingredients

Three types you'll master.

everything in the ontology is one of these ↓
01 · nouns
Object Types.

The things in your business — Customer, Order, Contract, Aircraft, Patient. Each type has a primary key, a title property, and a bunch of named properties backed by real datasets.

primary key title prop properties
02 · verbs
Link Types.

The relationships between objects — Customer places Order, Order contains Product, Patient seen-by Doctor. One-to-many, many-to-many, self-referencing — Foundry handles them all.

1-to-many many-to-many self-ref
GO!
03 · write-backs
Action Types.

The things you can do to the ontology — create, edit, delete, or run a side-effect. Every button in Workshop and every AIP agent decision flows through these, governed, validated, and audited by default.

create edit side-effects
🥞 the stack

How the ontology fits in Foundry.

three layers, one story ↓
— top layer · what people see
Consumers & Apps.
this is where the ontology becomes value
Workshop apps AIP · Logic + Agents OSDK · Ontology SDK Slate · dashboards Marketplace
— the semantic layer
Ontology — the shared model.
one source of business meaning · governed & live
Object Types Link Types Action Types Function-backed properties RBAC · audit · lineage
— foundation · raw material
Datasets & Pipelines.
the plumbing — data connectors, pipeline builder, code repos
Data Connectors Pipeline Builder Code Repositories Data Health
🌍 real deployments

Ontologies in the wild.

a few industries we'll draw examples from ↓
🏥
Healthcare
Patient · Encounter · Diagnosis · Prescription. Linking clinical events across facilities.
💰
Financial Services
Customer · Account · Transaction · Risk Alert. Real-time write-back to case workers.
📡
Telecommunications
Subscriber · Cell Tower · Outage · Ticket. Ontology-backed AIP agents in the ops room.
✈️
Aviation
Aircraft · Flight · Maintenance · Part. Traceable actions and full audit lineage.
🎁 you'll walk out with

Four things you can use Monday.

not slogans — actual decision frameworks ↓
Model a real domain in Foundry.
Turn a business description into Object Types, Link Types, and Action Types — a repeatable decision framework.
Design links that don't fall over.
Cardinality choices, backing datasets, and the traps that only surface at a few million nodes.
Wire actions into Workshop & AIP.
Turn a button click or an agent decision into a governed, auditable ontology write-back — parameter validation baked in.
Know what NOT to put in the ontology.
High-cardinality streams, blobs, transactional systems — where the ontology is the wrong home and what to do instead.
🎨 your presenter

Meet Chaitanya.

Chaitanya Vullaganti
Data & Ontology Engineer · Palantir Foundry

Chaitanya works day-to-day on real Palantir Foundry deployments — designing ontologies, wiring data connectors and pipelines, building Workshop apps, and shipping AIP-powered workflows. Her focus is the practical craft of turning raw datasets into a semantic layer the business can actually use, and the small design decisions that separate an ontology that scales from one that gets rewritten every six months. She'll be walking us through the working details in this session.

Save your seat.

free · 45 min + Q&A · Zoom invite instantly

register at the top ↑ Sat · Aug 15 · 11:30 AM EDT · 9:00 PM IST · sponsored by Think Big Techies
© 2026 Think Big Techies · thinkbigtechies@gmail.com · webinar series