Media and ad-tech data teams: less upkeep, more data products
Less upkeep, more data products. Most of the week on media and ad-tech data teams goes to upkeep: broken feeds, reruns, access queues and documentation. Data Workers hands that work to agents so the team ships the data products the business has been asking for.
Key takeaways
- •Who this is for. Data leaders and teams at streamers, media groups, ad platforms and ad tech.
- •The problem today. Most of the data team's week goes to overnight incident response, monthly reconciliation and clean-room tickets.
- •What changes. Specialised agents take that work over across the systems you already run: they catch bad data where it enters, fix it where it lives and verify the number before anyone reads it.
- •What stays with your people. Anything regulated or customer-facing waits for a named approver with the full diff, blast radius and rollback plan in front of them.
- •The result. Data teams that spend their time on new measurement and audience products.
The PRD for agentic data in media and advertising
This is the requirements document we would write with you. Everything below expands on it.
| Who | Data leaders and teams at streamers, media groups, ad platforms and ad tech. |
| The job | Measurement, audiences and viewership that tie out. |
| Environment | Kafka or Kinesis impression streams; Snowflake, BigQuery or Databricks; dbt; Airflow; AWS Clean Rooms, Snowflake or LiveRamp clean rooms; billing systems, and the other platforms your teams run. |
| Constraints | Regulated and customer-facing numbers always go to a named approver; agents run in your infrastructure under the roles you already govern. |
| Success is measured by | Delivered vs billed reconciled: from about a month to 8-15 min; Account manager question: from about a day to seconds; Clean-room access for an advertiser: from about a week to 5 min; sandbox and cost cleanup closed autonomously 88% of the time. |
| Owned by | The head of data or CDO, with domain owners as named approvers. |
| Pilot scope | Delivered-vs-billed at L2. |
What going fully agentic and autonomous means in media and advertising
- •Put your data pipelines on autopilot. Agents catch bad data where it enters, fix it where it lives and verify the number before anyone reads it.
- •Run the data back office without the busywork. The week that goes to overnight incident response, monthly reconciliation and clean-room tickets moves to agents.
- •Scale data without scaling headcount. New sources, partners and data products are onboarded as missions, not projects.
- •Get the plumbing off your plate. Data teams spend their time on new measurement and audience products.
The shift
Media companies used to sell inventory. Now they sell data products: first-party audiences, clean-room measurement, reach and frequency, viewership across every app and brand. Every major media group has built an identity spine and an audience platform in the last few years.
The data behind those products is some of the fastest-changing in any industry. Ad servers ship releases weekly, apps change how they log playback, identity providers change formats and measurement vendors change methods. Each change breaks a number an advertiser pays for. And inside a media group, ad sales, direct-to-consumer, programming and finance each count a view differently.
Our view: every number a media company sells should be verified by agents before anyone sees it. Agents keep event sources, identity and measurement data in step, reconcile delivered against billed every day, hold one definition of a view across every org, and fix what breaks overnight.
How data work in media got here
The ratings era. Panels and ratings were the currency. Data teams reported them.
The first-party data era. Streaming, ad-supported tiers and the end of third-party cookies pushed media groups to build identity spines, audience graphs and clean rooms. Data teams became the product teams behind ad sales.
The agentic back office. Agents run the upkeep and verification behind those products so ads and audience teams ship new measurement and audience products.
The three eras side by side:
Each era changed the tools. Only the agentic back office changes who does the work.
A day on an ads data team today
At 2am an ad server release renames a field. Two hourly partitions load with zero viewable impressions. The 9am delivery report goes out short to 40 advertisers. By noon, three account managers are discussing makegoods. A data engineer spends the day tracing it and the next day backfilling.
That morning is typical, not exceptional. Across a week, this is where the team's time goes, today and with agents on the back office:
Time spent on new data products moves from 16% of the week to 50%. Most of the rest becomes review: people steer and approve agent work instead of doing it by hand.
Then and now
| Work | Today | With Data Workers |
|---|---|---|
| An ad server or app changes an event | Found when a report goes out short | Detected on the first batch, mapped, rerun and verified overnight |
| Delivered vs billed reconciliation | By hand at month end | Daily, with one decision escalated when they diverge |
| Definitions of a view, impression and reach | Different in every org | One governed definition everywhere |
| Clean-room access for new advertisers | A week of tickets | Policy-driven, same day, every grant logged |
| Identity match-rate drops | Found when lift studies lose power | Monitored per provider, traced on the first drop |
| Event-table cost | Petabytes nobody dares delete | Retired once lineage proves nothing depends on it |
These are the impact metrics we move:
- •Delivered vs billed reconciled: from about a month to 8-15 min
- •Account manager question: from about a day to seconds
- •Clean-room access for an advertiser: from about a week to 5 min
Across the routine work:
The biggest change is Delivered vs billed reconciled: from about a month to 8-15 min. Work that waited in queues finishes in minutes or hours, because agents detect, fix and verify in one pass.
Use cases, in depth
1. The 2am field rename
Today. Found in the morning report. Fixed over two days. Makegood conversations start.
With Data Workers. Real-Time Streaming detects the new schema, Schema Evolution proves it is a rename, Pipeline Building fixes the models, Incident Debugging reruns the bad partitions and Conductor verifies delivery ties to billing, holding the corrected advertiser report for one approval. Read the full walkthrough.
What changes. The 9am report is right.
2. One definition of a view across the media group
Today. DTC, ad sales, programming and finance report different numbers for the same title.
With Data Workers. Data Context Wizard holds one governed definition of a view, a viewer and a reached household. Every report, model and assistant reads it. When the definition changes, every consumer moves together.
What changes. The weekly viewership number stops being a negotiation.
3. Playback against vendor measurement, every day
Today. Discrepancies between streaming playback and vendor measurement surface in quarterly reviews.
With Data Workers. Conductor reconciles playback against vendor numbers daily, Quality Monitoring tracks drift per app and platform, and Incident Debugging traces divergence to a logging change, a vendor method change or real behavior.
What changes. Teams know which it is the next morning.
4. Training data for in-house GenAI
Today. Media and creative companies training their own models track rights to assets in spreadsheets.
With Data Workers. Access & Governance enforces license scope on every dataset and training job, and Context Wizard links each model version to the assets it trained on. See the GenAI training data playbook.
What changes. When a license changes, the team knows in minutes which models are affected.
Agents working together: The ad server renamed a field at 2am
The 2am field rename, end to end. Six agents and one approver fix it overnight. The ads data mart playbook has the full minute-by-minute walkthrough.
Minute by minute
This is an illustrative example. Impression logs stream from the ad server through Kafka into Snowflake; dbt builds delivery, reach and billing tables; the advertiser delivery report sends at 09:00. Advertiser-facing reports are at L2.
| Time | Who | What happens |
|---|---|---|
| 02:04 | Ad server | Release renames viewable_flag to is_viewable |
| 02:05 | Kafka | New schema version on the impression topic |
| 02:06 | Schema Evolution agent | detect_schema_change; value distribution proves a rename |
| 02:11 | Pipeline Building agent | Model diff with tests; staging merged at L3 |
| 02:20 | Incident Debugging agent | execute_backfill on two hourly partitions |
| 02:41 | Quality Monitoring agent | Delivery ties to billing; report change held at L2; create_jira_sm_ticket |
| 07:52 | Spellbook Data Catalog | Measurement lead approves |
| 09:00 | Report send | 40 advertisers receive correct delivery |
How it fits your stack
Data Workers connects to what you already run: Kafka or Kinesis impression streams; Snowflake, BigQuery or Databricks; dbt; Airflow; AWS Clean Rooms, Snowflake or LiveRamp clean rooms; billing systems. Each agent is an MCP server, so the same agents work inside Claude Code, Cursor or Codex and in Slack or Teams through Spellbook. Nothing migrates.
- •Runs in your infrastructure with your credentials and your model key. The hosted Autonomous Data-Conductor sees workflow metadata only, never rows.
- •Uses the access you already govern. Agents act under scoped service identities and your existing roles; every grant goes through provision_access and is logged.
- •Proposes changes as diffs in your dbt project, orchestrator and warehouse, with blast radius, tests and a rollback point, and applies them only at the autonomy level you set for that domain.
- •Open-source core. The Apache-2.0 community edition ships 11 agents and 160+ MCP tools, so your engineers can read exactly what the agents call.
How autonomy expands over the first year
Autonomy is earned per task class. In a mature deployment, it looks like this:
Sandbox and cost cleanup closes autonomously 88% of the time. For Advertiser-facing reports, autonomous closure is 0% by design: every change is proposed with a full diff and approved by a named person.
A typical path to get there:
- •Month 1, L1. Agents map every event source, identity input, metric definition and lineage path.
- •Months 2 to 3. Sandbox cleanup, cost fixes and catalog upkeep at L4. Reruns, backfills, new field profiling and quality rules at L3.
- •Months 4 to 6. Daily delivered-vs-billed and playback-vs-vendor routines. New measurement products built as missions.
- •Months 6 to 12. Repair classes with a clean record move up. Advertiser-facing reports, billing and sellable audiences stay at L2. Makegoods and advertiser communications stay with people.
The same path as a curve, month by month:
By month 12, more than 46% of agent actions run fully autonomously and manual work falls below 10%, while advertiser-facing changes stay with a named approver.
What happens when an agent is wrong
- •At L2 propose, nothing changes until a named person approves. The proposal shows the full diff, the blast radius and how to roll it back. Unanswered requests expire and escalate.
- •At L3 act, reversibly, every change keeps a rollback point, and Conductor verifies the downstream number after the change. If verification fails, the change is reversed and the domain drops a level.
- •No agent approves or promotes its own work. Autonomy only grows on evidence, per repair class, and repeated errors shrink scope automatically.
- •One org-wide stop halts all autonomous dispatch, and every action lands in a hash-chained audit trail you can export with generate_audit_report.
Governance and regulation
Ad and viewership data carry consent and personal-data obligations. Data Workers classifies personal data, scopes clean-room access and enforces consent per brand and region.
Data Workers doesn't make you compliant; your controls, people and assessors do. It supplies controls and evidence wherever agents touch data. Not legal advice.
The alternatives media and advertising data teams weigh
Build it yourselves. A strong team can wire a coding agent to MCP servers for the warehouse and dbt. The work is everything around it: routing approvals to the right owner, blast radius, rollback, receipts, and current context about every other system.
Observability, catalogs and point tools. Observability alerts, catalogs document, and quality tools test. Data Workers takes their signals, diagnoses, proposes the fix, verifies it and leaves the receipt.
Platform-native agents. Snowflake made its managed MCP server and Cortex Agents generally available on Nov 4, 2025, and Databricks made Genie One generally available on Jun 16, 2026. They are excellent inside their platform; this work crosses the source systems, orchestrator, warehouse and BI, and Data Workers works across all of them.
In this space specifically. Ad servers, clean rooms and measurement vendors each own a slice; Data Workers ties them together and to billing.
The case for your CTO
What this replaces. The week spent on overnight incident response, monthly reconciliation and clean-room tickets.
What your teams get back. Time for new measurement and audience products, on data that is verified before anyone reads it.
Integration effort. Connecting is configuration, not a migration. The agents attach over MCP to the platforms you already run, under the roles you already govern, inside your own infrastructure.
Risk and control. Autonomy is set per domain, starts at L1 and rises only on evidence. Regulated and customer-facing numbers stay at L2 with a named approver, and one org-wide stop halts all autonomous work.
Why now. Ad-supported streaming and clean rooms multiplied volume.
The first win. Delivered-vs-billed at L2. The pilot below lays out the scope, what you see each week and the cost.
The climb: from assisted to autonomous to agentic
Our thesis is that the data platform climbs the way analytics did, in stages. In media and advertising, that climb looks like this:
- •Assisted. Agents work alongside each data team, read the whole stack and propose every fix with a receipt. Regulated and customer-facing numbers always go to a named approver.
- •Autonomous, under your governance. The upkeep behind overnight incident response, monthly reconciliation and clean-room tickets runs on its own, domain by domain, as the receipts earn it.
- •The agentic enterprise. The platform largely runs itself. People set direction and standards, and the teams spend their time on new measurement and audience products.
You choose the altitude, every step is visible and reversible, and the pilot below is the first rung.
The pilot: how media and advertising data teams start
The pilot is how a team proves the value on its own data before committing.
- •What we connect. Kafka or Kinesis impression streams; Snowflake, BigQuery or Databricks; dbt; Airflow; AWS Clean Rooms, Snowflake or LiveRamp clean rooms; billing systems, over MCP, inside your infrastructure. Nothing migrates.
- •Scope. Delivered-vs-billed at L2.
- •First week. Agents run at L1 and show the fixes they would have made, with receipts.
- •By week two. Reversible upkeep (reruns, staging fixes, documentation, cleanup) runs on its own; regulated and customer-facing numbers stay at L2 with a named approver.
- •At day 30. We review the impact metrics against your baseline:
- •Delivered vs billed reconciled: from about a month to 8-15 min
- •Account manager question: from about a day to seconds
- •Clean-room access for an advertiser: from about a week to 5 min
- •the share of sandbox and cost cleanup closed autonomously, targeting 88%
- •What you decide. Which domain or team goes next. The pilot fee is credited in full against the first year.
- •Cost. Start with a pilot ($7,500 one-time, credited in full against the first year). Scale is from $1,000 a month and Enterprise from $3,000 a month, billed annually, with unlimited seats, no usage meter and no markup on model spend. See pricing.
Teams we work with in media and advertising
FAQ
Clean room? Access and PII around it.
Streaming? Yes.
Who approves advertiser reports? A named lead at L2.
Where do the agents run? In your infrastructure, with your credentials and model key. The hosted Conductor sees workflow metadata only.
Do we have to migrate anything? No. Your warehouse, dbt project, orchestrator, catalog and BI stay where they are.
Sources
- •Disney Audience Graph, VentureBeat: https://venturebeat.com/2022/02/18/how-disney-built-its-audience-graph-to-meet-data-demands
- •Data Workers pricing: https://dataworkers.io/pricing/
- •Data Workers open-source repository (tool registrations): https://github.com/DataWorkersProject/dataworkers-claw-community
- •Snowflake, managed MCP server and Cortex Agents GA release note (Nov 4, 2025): https://docs.snowflake.com/en/release-notes/2025/other/2025-11-04-cortex-agents-mcp
- •Databricks, Genie One launch (Jun 16, 2026): https://www.databricks.com/company/newsroom/press-releases/databricks-launches-genie-one-all-new-agentic-coworker-every-team