01 / 06
Connect
Prebuilt adapters for the apps you already run, and custom ones in days to weeks.
Petrel StudioKingdomGeoGraphixPetraOpenWorksCompass / EDMIHS DataIn-House Databases
One Well. One Answer. Everywhere.
ODIN connects to the systems you already run, compares the same well across all of them, and corrects it according to rules you set. It runs continuously, so the picture is always current.
02 / 06
Bring It Together
Every source lands in one place, each well matched across systems and tagged by your rules.
Kestrel 14H · Wellbore · 6 sources
03 / 06
Check
Three levels of quality, checked continuously, with subsurface logic built in.
CompletenessEvery required field filled?Pass
ConsistencySame value everywhere?Pass
ValidityTD matches the survey?Flagged
04 / 06
Decide
Each issue is fixed automatically, sent for approval or flagged for review.
Kingdom · TD
05 / 06
Correct
Fixes flow back to the systems you choose, with a full audit trail.
SYNCEDTD 12,480 ft → 6 systems · logged
06 / 06
Keep Watching
New data and edits are checked as they arrive, and the loop starts again.
WATCHING4,812 wells · 6 systems
CAUGHTKestrel 15H · TD edited in Petra → back to Check
On-prem, cloud or hybridMinimal IT after setupTurn-key, not a dev kit
What Changes
Without ODIN
With ODIN
Reactive Fix it after it breaks
Continuous Data flows across all systems
Scattered Logic spread across scripts and teams
Centralized Rules and logic, maintained in one place
Hidden Data issues compound silently
Surfaced Flagged in near real time, with old and new values
Inconsistent Systems disagree, and trust erodes
Aligned Systems you can trust
Unicorn-Dependent One specialist keeps it running
Easy to Manage Simple to hand off, no unicorn needed
Now available
Staging
A Layer for the Best Answer
Sometimes no single system has the whole answer. The surface location is best in one source, the total depth in another, and the status in a third. And sometimes the system of record is slow to update, so a correction has nowhere to go yet.
ODIN's staging layer solves both. It holds the best answer for every well, blended attribute by attribute from all your sources, based on rules you set for which source wins.
Sources
- Petrel Studio
- Kingdom
- IHS
- Corporate DB
Staging
1Well A-12 · Best Answer
Total Depth
12,480 ft
Surface Location
31.84217, −102.3741
Status
Producing
Applications3
- Petrel StudioSynced
- KingdomSynced
- OpenWorksSynced
- Corporate DB2Queued
- 1
A Blended Master
The most trusted value for each attribute, even when it comes from different sources.
- 2
A Place for Corrections to Wait
Approved fixes can be saved and sent out, even before the corporate or vendor system catches up.
- 3
One Answer to Share
The staging layer feeds your applications, and your workflows control exactly where it goes.
You Decide What Matters, and How Data Flows
Automation is only useful if you trust it. Setup is configuration, not coding. Your team answers a few questions, and ODIN applies the answers everywhere.
Epics
Start with the data you care about.
Instead of choosing rules, you choose the data that matters, like surface locations on wells, and mark it Critical or Important. ODIN turns on the right checks automatically.
Start with one epic in one area, watch quality improve, then add the next. Data quality becomes a journey taken one bite at a time.
EpicSurface Locations
- You choose the dataWellsSurface LatitudeSurface LongitudeSurface CRS
- You set the importanceCriticalImportant
- ODIN turns on the checks
- Completeness
- Consistency
- Validity
- Integrity
- Correction
Next epicsDeviation Surveys · Marker Sets
Workflows
Your governance applied consistently.
Classifications sort wells into lanes, and each lane follows its own rules for every kind of change.
Critical changeWell A-12 · TD 12,455 → 12,480 ft
- Company OperatedA Person Chooses
- Company Interest, Not OperatedAutomatic, With Approval
- Public WellsAutomatic
Example only. Your team sets the lanes and actions, and a scenario tester shows exactly what will happen before anything is written.
What Comes Built In
Most data tools hand you a blank canvas. ODIN arrives already knowing subsurface data and how to manage it.
Checks Built From Experience
Data can look complete and still be wrong. ODIN checks three levels deep.
- 01
CompletenessIs it there?
Required fields populated in every connected system.
Most tools stop here
- 02
ConsistencyDoes it match?
The same answer in OpenWorks, Petrel and every other source.
- 03
ValidityDoes it make sense?
Where ODIN's domain knowledge earns its keep.
Validity in Practice
One well, three problems most tools never see.
- 1
Wellbore and survey disagree on depth. TD is 12,480 ft, the last survey point 12,385 ft. Each possible cause has its own fix.
- 2
Data deeper than the survey. Perforations below the last survey point drop out of most displays.
- 3
Two wellpaths for one well. ODIN recalculates the path from the survey and corrects the stored one.
Rules
A library of checks for every data type. Choose the data, and the right rules switch on.
Classifications
Wells sorted by business context, each class following its own workflow.
Adapters
Pre-built for the applications you run, and tested against new versions first.
Knowledge of the Data
Native models for subsurface data, with CRS, datum and unit handling.
Knowledge of the Process
Guided setup, visual workflows, and a record of every read, write and fix.
A Team That Knows It
An ODIN Solutions Engineer with decades of experience at operators.
Supported Data Types
WellsWellboresDeviation SurveysPerforationsProduction DataMarker TopsLog CurvesWell ZonesCasingsLinersTubingsDrilling Muds
We're constantly growing our library of data types and adapters.
A Foundation for AI
AI is only as good as the data it can access, consume, and understand.
AI-Ready Data
Clean, consistent, connected subsurface data, with a record of where every value came from, for AI agents and machine learning models.
MCP Server
Connects ODIN's trusted data to AI assistants and copilots through the Model Context Protocol, an open standard for linking enterprise data to AI.
Well A-12 · Total DepthTrusted
12,480 ft
ODIN MCP ServerModel Context Protocol
AI AssistantsCopilotsAI AgentsML Models
Questions We Hear Before Every Project
How much of our team's time will this take?
Initial
Data Specialist 4–8 hours/week. Business User engagement 1 hour/week.
Ongoing
Typically a few hours a week for data exceptions. IT's role is basic server support.
We've been burned by tools before. Why is this different?
We don't ask you to take our word for it. We prove ODIN on your data, in your environment, before you commit.
Will our users actually use it?
ODIN works in the applications they already use, and success is defined up front with the people who use the data every day and know what good looks like.
Why not build it ourselves?
You could. The question is whether subsurface data integration is where your best people should spend their time, and who maintains it when they move on. And would it give your users the same visibility into every project, side-by-side comparisons across sources, a full change history, and the data exceptions that need a decision?
Do we need to write code or hire a specialist to run it?
No. Setup is configuration, not coding. Rules and adapters come pre-built. Your team answers a few questions about what matters and how data should flow. ODIN applies those answers everywhere.
Where does it run, and what does IT need to do?
In your environment: on-premises, in the cloud, or both. Your data stays with you.
How We Work With You
Start small, prove it on your own data, and grow at your own pace.
- 01ConversationA few calls
We learn your systems, the data that matters most, and where it hurts. Then we walk you through ODIN.
- 02Environment PrepAbout 15–30 days
Your team sets up access to the sources ODIN will work with. We guide you through exactly what's needed.
- 03Proof of ConceptAbout 30 days
ODIN runs on your data, in your environment, against success criteria we define together.
- 04We StayOngoing
We maintain ODIN, monitor it, and handle new edge cases as they surface. New data types and areas are added when you're ready.
See it on your own data.
Every engagement starts with a conversation, then a proof of concept in your environment.