Manage Data Where It Lives With Dat­ab­ricks Apps

As orga­ni­sa­ti­ons scale their data estates, more and more of the day-to-day work shifts from buil­ding new pipe­lines to mana­ging, main­tai­ning, and gover­ning the ones that alre­ady exist. Someone has to check that yesterday’s load ran cle­anly, that a schema change did­n’t sil­ently break a down­stream report. Dat­ab­ricks Apps close that gap. They address this by pro­vi­ding light­weight, gover­ned appli­ca­ti­ons that run directly inside the Dat­ab­ricks workspace. They turn data pro­ducts and models into inter­ac­tive tools wit­hout export­ing data to exter­nal sys­tems while also kee­ping gover­nance guar­drails at a high level.

Dat­ab­ricks apps are desi­gned for ope­ra­tio­nal work­flows: moni­to­ring pipe­line health, vali­da­ting data qua­lity, hand­ling schema chan­ges, coor­di­na­ting appr­ovals, and pro­vi­ding safe inter­faces for busi­ness users to inter­act with data. They are espe­ci­ally useful for self‑service tasks, such as cura­ted access and appr­oval flows tied to gover­nance con­trols. These apps pro­vide a gover­ned, in-workspace inter­face to per­form  ope­ra­tio­nal tasks. Teams can also use Dat­ab­ricks Apps to moni­tor pipe­line health, run self-ser­vice data qua­lity checks, detect anoma­lies, track cost and usage, sup­port meta­data ste­ward­ship, and give busi­ness users safe, cura­ted access to data, all while pre­ser­ving the secu­rity, lineage, and per­mis­si­ons the plat­form alre­ady provides.

Table of contents

Esti­ma­ted rea­ding time: 5 Minuten

What Dat­ab­ricks Apps Are and Why They Matter

Dat­ab­ricks Apps are web appli­ca­ti­ons that are deployed directly to the Dat­ab­ricks envi­ron­ment. They allow teams to trans­form repeata­ble data work­flows like gover­nance decis­i­ons, qua­lity gates, and ope­ra­tio­nal tasks into user‑friendly inter­faces that mini­mise manual inter­ven­ti­ons and reduce risk. Ins­tead of ad‑hoc note­books or SQL queries, you can pro­vide struc­tu­red “front doors” for busi­ness users, data ste­wards, and engi­neers. A way to do con­trol­led appr­oval flows, gover­ned data edits, and self‑service ope­ra­ti­ons that fully respect Unity Cata­log per­mis­si­ons and auditability.

Dat­ab­ricks posi­ti­ons Apps as a way to quickly build secure data and AI appli­ca­ti­ons using fami­liar Python frame­works like Dash while kee­ping deploy­ment and main­ten­ance over­head mini­mal. For Dat­ab­ricks, they are the natu­ral next step towards an end-to-end data plat­form that covers every step from data inges­tion over trans­for­ma­ti­ons, data qua­lity and gover­nance up to BI and operations.

What These Apps Look Like in Practice 

Apps serve as con­trol­led, gover­ned inter­faces for pro­ces­ses that would other­wise require direct data access or infor­mal appr­ovals. Com­mon examp­les include: 

  • Appr­oval work­flows for gover­nance decisions: 
    • Self-ser­vice access con­trol portal
    • PII (Per­so­nally Iden­ti­fia­ble Infor­ma­tion) data clas­si­fi­ca­tion, tag­ging and anony­mi­sa­tion based on engi­nee­ring, busi­ness and GDPR requirements
  • Qua­lity gate dash­boards that vali­date data before promotion
  • Self‑service ope­ra­tio­nal tools enab­ling busi­ness users to inter­act with cura­ted pro­ces­ses wit­hout using note­books or SQL 
  • Gover­nance tool­kits like Ontos, a data con­tract and data pro­duct gover­nance plat­form built and main­tai­ned by Dat­ab­ricks, which fol­lows the open source ODPS and ODCS stan­dards to unify how busi­nesses buy, manage, and scale data products.

A Key Archi­tec­tu­ral Detail: The Read/Write Access Pat­tern 

Effec­tive gover­nance hin­ges on how the app inter­acts with data. Dat­ab­ricks Apps allow to imple­ment a hybrid access pattern:

Both the access to the app and the read ope­ra­ti­ons inside the app occur “on behalf of” the authen­ti­ca­ted user lever­aging their indi­vi­dual access. Unity Cata­log per­mis­si­ons apply auto­ma­ti­cally, ensu­ring data access is con­sis­tent and traceable. This also ensu­res that data mas­king and row level secu­rity rules work con­sis­t­ently accross the platform.

Wri­tes are per­for­med via a dedi­ca­ted ser­vice prin­ci­pal, allo­wing updates only through the app’s vali­da­ted busi­ness logic, while recor­ding who initia­ted the change for full auditability. 

A few Examples

Let’s explore some examp­les of how Dat­ab­ricks Apps can enhance com­mon data ope­ra­tion pat­terns that would other­wise require ele­va­ted pri­vi­le­ges and addi­tio­nal audit trails.

Manual Data Adjus­t­ments with Appr­oval Process

Whe­re­ver appr­oval pro­ces­ses or manual adjus­t­ments of data are nee­ded, apps can pro­vide a plat­form that pro­vi­des an inter­face for non-tech­ni­cal users.

In this exam­ple, an app is pro­vi­ded that expo­ses the func­tion­a­lity to manu­ally adjust KPI fac­tors applied during data pro­ces­sing. These adjus­t­ment fac­tors are other­wise pro­vi­ded by CSV or Excel. And while manual file trans­fer works, it is error-prone, espe­ci­ally due to loca­li­sa­tion and user error.

1. Request tab 
The reques­ter sel­ects an attri­bute, loads refe­rence data into a grid, edits only per­mit­ted fields, pre­views chan­ges, and sub­mits a “Change Request”. The sys­tem records all rele­vant meta­data (requested_by, requested_on, comm­ents) asso­cia­ted with the request. 
Only columns con­fi­gu­red in a YAML con­fig are ope­ned for chan­ges. If chan­ges are reques­ted and appro­ved suc­cessfully they flow back into the ETL pipe­lines based on their ori­gi­nal keys. The changable columns are high­ligh­ted in the UI.

2. Bulk Upload tab
The user can upload a full CSV or Excel file to manu­ally popu­late the plat­form with data. The data struc­ture and com­pli­ance with the expec­ted schema are che­cked. Within the same page the user instantly recei­ves feed­back on any issues.

3. My Requests tab 
Reques­ters can track each sub­mis­sion, with sta­tus deri­ved from appr­oval metadata. 

4. Approve tab 
Appro­vers view only requests sub­mit­ted by others, enfor­cing the four‑eyes prin­ci­ple. Appr­ovals cap­ture appro­ver meta­data and may trig­ger down­stream pro­ces­sing or noti­fi­ca­ti­ons. Cur­rent and pro­po­sed ent­ries are dis­played next to each other for visi­bi­lity.
Appro­vers can sel­ect indi­vi­dual rows to approve. On appr­oval repro­ces­sing of the down­stream pipe­line starts automatically.

PII Data Clas­si­fi­ca­tion, Tag­ging and Anony­mi­sa­tion App

Strong data gover­nance plays a key role in GDPR and PII and ties in clo­sely with a busi­ness‘ data stra­tegy. Espe­ci­ally the dis­co­very and clas­si­fi­ca­tion of such data are tedious and often a manual process.

This exam­ple is ano­ther appr­oval work­flow app. It stan­dar­di­ses how orga­ni­sa­ti­ons iden­tify, vali­date, and approve sen­si­tive data in Dat­ab­ricks Unity Cata­log. The app ensu­res that data is tag­ged based on com­pli­ance needs before mas­king, hash­ing, and rela­ted rules are applied. Admi­nis­tra­tors con­fi­gure appro­vers on a pro­ject level while review­ers pro­cess assi­gned appr­oval cases via a gui­ded work­flow. The work­flow cap­tures table con­text and enforces column-level clas­si­fi­ca­ti­ons, cate­go­ries, and anony­mi­sa­tion set­tings. Built-in guar­drails pre­vent inva­lid classification/category com­bi­na­ti­ons and sup­port dis­co­very of miss­ing tables or columns. Fina­li­zing the pro­cess applies gover­ned tags to tables and columns. A gover­ned Tag is an account-level tag, set up by admi­nis­tra­tors. They can enforce poli­cies or cus­tom rules for attri­bute based access con­trol or flag data based on GDPR.

Dat­ab­ricks Labs‘ Ontos

Ontos is a Dat­ab­ricks Labs pro­ject that adds a busi­ness gover­nance and seman­tic layer on top of Unity Cata­log. It’s goal is to help orga­niza­ti­ons move bey­ond tech­ni­cal meta­data (tables, sche­mas, lineage, per­mis­si­ons) and manage the busi­ness mea­ning, owner­ship, qua­lity, and life­cy­cle of data products.

Built in React, directly inte­gra­ted with Dat­ab­ricks Unity Cata­log and Dat­ab­ricks Lake­base as a fast sto­rage layer it offers a stack of data gover­nance functionality:

  • AI-assis­ted data discovery
  • Data pro­duct management
  • Busi­ness glossaries
  • Data con­tracts
  • Semantic/ontology manage­ment
  • Ste­ward­ship workflows
  • Data mar­ket­place capabilities
Dat­ab­ricks Labs‘ Ontos – Gover­nance Suite – Landing Page

Bot­tom Line 

Dat­ab­ricks Apps extend the Lake­house bey­ond ana­ly­tics by brin­ging gover­ned ope­ra­tio­nal pro­ces­ses directly into the Dat­ab­ricks plat­form. They enable orga­niza­ti­ons to build secure, user-fri­endly inter­faces for appr­ovals, data ste­ward­ship, qua­lity manage­ment, access requests, and gover­nance work­flows wit­hout moving data out­side the plat­form. By com­bi­ning user-based read access with con­trol­led write ope­ra­ti­ons, Dat­ab­ricks Apps help enforce com­pli­ance, improve audi­ta­bi­lity, and reduce manual effort while pre­ser­ving Unity Cata­log secu­rity, lineage, and gover­nance stan­dards. Whe­ther sup­port­ing appr­oval pro­ces­ses, PII clas­si­fi­ca­tion, or broa­der data pro­duct gover­nance through solu­ti­ons such as Ontos, Dat­ab­ricks Apps pro­vide a prac­ti­cal way to turn gover­nance from a manual bur­den into an inte­gra­ted, sca­lable ope­ra­ting model.