SEMORA IS LIVE
First 10 users to register get dedicated 1V1 technical support and
priority response—from data connection to trusted answers.
· INTALINK FOR DB · VS CODE EXTENSION
Less JOIN guesswork.
Less rework.
More time to model.
You know SQL. Finding the right relationship should not keep holding you up.
Searching tables, retracing models, asking colleagues and debugging inflated totals? IntaLink for dbt brings candidate JOINs, relationship evidence and some fanout warnings into VS Code—so you have something concrete to check while you model.
IntaLink Cloud + a VS Code extension. Built for your existing dbt project.
✅ Evidence you can inspect
✅ You decide what to use
Less hunting for JOIN keys
See candidate tables, columns and evidence where you write SQL.
Earlier clues to double-counting
Review some join risks in code, not only when a report fails to reconcile.
Reuse what the team has reviewed
Share confirmed or no-longer-recommended statuses, and preview dbt tests.
LESS INVESTIGATION / MORE MODELING
Building models?
Or still hunting for relationships?
Taking over a project, changing a model or delivering a report: more SQL is not the whole answer. You need to know what to join, why it fits and where measures might be repeated.
What interrupts your work
Without relationship evidence
WITH INTALINK FOR DBT
01 FIND THE JOIN
Which table? Which key?
Where do I start?
Browse schemas, read old SQL and ask a colleague to find a plausible relationship.
02 TRACE THE SOURCE
The column was renamed.
Where did it come from?
Follow model after model upstream to work out the original field and relationship.
03 CATCH THE FANOUT
The SQL ran.
Why is the total too high?
Wait for a reporting discrepancy, then investigate join keys, row counts and aggregation grain.
04 CHECK THE REVIEW STATUS
Has the team reviewed
this relationship?
Review status is hard to find. A new teammate has to check whether a relationship has already been confirmed.
Inspect candidate tables, columns and relationship graphs in VS Code. Review the evidence, then choose what to insert.
From guessing broadly → reviewing specific candidates
Trace renamed columns and type casts through multiple ref() layers. Calculated columns are not traced.
From repeated backtracking → a trace to review
See some potential JOIN double-counting risks and their explanations at the code location. Start with the join that needs review.
From debugging a wrong total → reviewing the risk in code
Mark relationships as confirmed or no longer recommended. Share those statuses in the project and preview dbt tests.
From repeated review → shared confirmation status
Reduce repetitive evidence gathering—not the engineer’s review and validation.
RUNS SUCCESSFULLY ≠ ADDS UP CORRECTLY
Two orders total 300.
Why does this JOIN return 400?
The SQL is valid, but an order amount repeats across line items. See the problem, then explore candidate joins, column traces, risk warnings and team review statuses.
WHAT CHANGES IN YOUR WORKDAY
Not just another tool.
Less repeated investigation.
From taking over a project to writing a model and handing it off: get concrete relationships, review clues and decisions your team can actually use.
01 / START A PROJECT
Take over a project
with a place to start.
Rely less on column-name guesses and scattered explanations. Inspect the graph, candidate keys and source trace before modeling.
What you get
Candidate tables and columns, source traces and reviewable evidence.
What to check
Can you find a useful candidate for an unfamiliar JOIN and explain why it fits?
02 / BUILD A MODEL
Review a risk before
handing over the total.
Look beyond whether a JOIN works to what it could duplicate. Review some risks while the model is being written.
What you get
A risk warning at the code location and an explanation to investigate.
What to check
Try a known fanout case and review whether the warning and explanation help.
03 / WORK AS A TEAM
Make the team’s
review status visible.
Show teammates which relationships are confirmed and which should no longer be recommended, without repeating the same review.
What you get
Confirmed or no-longer-recommended statuses, plus dbt tests to preview.
What to check
Confirm a relationship, inspect its shared status, then review and run a generated dbt test.
For your first trial, choose one JOIN you are unsure about.
Find a candidate, understand its evidence and decide whether to use it. That matters more than installing an extension.
CLOUD DISCOVERY / EDITOR EXPERIENCE
Discover in the cloud.
Use the evidence in your editor.
IntaLink for dbt combines IntaLink SaaS with a VS Code extension. Discover relationships first, then sync the results into your project.
01 / YOUR SOURCE
Your data source
Authorize a database connection and choose the tables and columns whose metadata and values can be analyzed.
Validated in MySQL / PostgreSQL test projects
→
02 / INTALINK CLOUD
IntaLink SaaS
Read table and column metadata, compare column values, and discover candidate relationships with supporting statistics and evidence.
Microsoft cloud SaaS · no self-hosted IntaLink service required
→
03 / YOUR EDITOR
VS Code + dbt
Sync results and read project structure to surface candidate JOINs, relationship graphs, evidence and some risk warnings.
Review → set a status → preview dbt tests
Not a dbt replacement or a standalone SQL generator
Keep your project and development workflow. IntaLink adds relationship context to help you make informed joins.
Rule-based extension checks. Optional AI review in the platform.
The extension’s own checks do not use AI models. IntaLink Cloud offers optional AI review. Engineers still validate business meaning, grain and results.
KNOW WHAT IS CONNECTED
Know what syncs.
Know when data is queried.
Understand the separate roles of the platform, extension and active verification before connecting your data.
Normal sync: relationships, not business rows
During normal sync, the extension receives relationship results and statistics, such as distinct-value counts—not business rows. It does not directly connect to your database for this sync.
Cloud discovery reads column values
IntaLink Cloud uses your authorized connection to read table and column metadata and compare column values. Discovery is not schema-only. Confirm access scope, processing, residency and retention before connecting.
Dedicated tokens: system secure storage
Create a token for the “VS Code extension” purpose. Tokens are stored in the computer’s system secure storage, not as plaintext in project files.
Active verification: initiated by you
When you choose “Verify with dbt”, your own connection queries the two columns’ minimum, maximum and counts. Results are displayed only on your computer. This is separate from normal sync.
Before onboarding, confirm permissions, data handling and retention for the deployed version. This demo page does not request database passwords or access tokens, or upload your dbt project.
FROM CLOUD TO YOUR FIRST USEFUL JOIN
From signup to your first
useful relationship suggestion.
01
Create your IntaLink Cloud account
Register and sign in to the cloud service; there is no platform to deploy yourself.
Have an authorized data source ready.
02
Discover and inspect relationships
Configure authorized connection. IntaLink reads table/column metadata, compares values to find candidates, then reviews one relationship and its evidence first.
Review its source, evidence and review state.
Bring VS Code, a dbt project and an authorized data source. After cloud discovery and extension setup, evaluate the suggestions against your own work.
03
Create your IntaLink Cloud account
Install IntaLink for dbt in VS Code, enter the platform address and a dedicated token, then sync results.
oken path: Settings → Account & security → Access tokens. Choose “VS Code extension”.
04
Open your dbt project for structure detection and relationship mapping. Run dbt parse if structure information has not been generated.
Inspect and use your first useful suggestion in a real development task.
Validated scopeVS Codedbt test projectsMySQL / PostgreSQLCheck other databases and version compatibility before use.
BEFORE YOU CONNECT
Clear answers.
Before you connect.
Start with your current project, and check capabilities, prerequisites and data boundaries.
YOUR PROJECT / YOUR FIRST USEFUL RELATIONSHIP
Start with the JOIN
you are least sure about.
Tell us what is slowing you down: finding JOINs, tracing columns, investigating fanout or sharing decisions. Confirm setup requirements, then test one concrete improvement in your own dbt project.
✅ Check version and source compatibility
✅ Connect the cloud and extension
✅ Evaluate suggestions against your own task