Opportunities › DAG control plane for warehouse pipelines › Proposed app
Tracewell
One control plane for pipeline lineage, masking, cost and replay across warehouses
- Score
- 47
- confidence 56
- Evidence
- 16
- items on the page
- Rivals
- 4
- named in sources
- First version
- 60 d
- estimate
- Type
- Desktop app
- PWA: runs in the browser, installs on Mac and Windows
- Market since
- 2018
- Established market · Verified
ScreensConcept
Desktop app: Data engineers work in dense dashboards and audit trails for hours, reviewing lineage graphs, masking policies, and cost analysis across warehouse models. Sample screens on a computer, in the proposed brand. Their numbers are sample data.
Change impact
Lineage compiled in 3s14 columns
About this appAI proposal
Tracewell sits on top of your warehouses and owns the pipeline graph. It reads your models, works out column lineage when the project compiles, and shows which columns are sensitive before anything runs. It works with Databricks, Snowflake, BigQuery and DuckDB, and does not replace them or your orchestrator.
Classification tags, per-environment masking policies, audit trails and cost per model are in one place instead of being spread over each warehouse's own tools. A CI gate fails a change that exposes a tagged column or breaks a budget.
Branch a pipeline, replay it against a past run, and compare results before merging. It is meant for data engineers who handle governed data in several environments and now check this by hand.
- Tagline
- Know every column, cost and policy before the pipeline runs.
- Voice
- PreciseCalmDirect
- Why this name
- Trace is what lineage does, and well suggests a reliable source. Two plain syllable groups, easy to say and spell.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
Compile-time column lineage
Lineage is computed when the project compiles, not by crawling the warehouse afterwards. Each pull request shows exactly which columns it touches.
No member yet
Classification that follows the column
Tag a column once as PII or financial. The tag carries through every downstream model that uses it.
No member yet
Masking policies per environment
Define one policy and apply it to Databricks, Snowflake and BigQuery. Dev can see masked values while prod stays restricted.
No member yet
Cost per model
Attribute scan and compute cost to each model and run. Set budgets and get warned before a change overshoots.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#1F5C4DDeep Pine · primary#F6F3ECPaper · surface#1F2A28Ink · text#D9822BAmber · accent#3B6EA5Slate Blue · data#B23A48Brick · alert
Typography
Aa
Tracewell
PricingAI proposal
Verified pages show Dagster at $10 to $100 a month and dbt at $100 a month plus a free plan. Team is set near that range. The material gives no price for governance; all tiers are a starting point to test.
Starter
Free
- 1 warehouse
- Compile-time lineage
- Up to 50 models
- Community support
Team
$100 a month
- Up to 3 warehouses
- Masking and classification
- Cost per model and budgets
- CI gates
- Audit log, 90 days
Enterprise
Contact us
- Unlimited warehouses
- Audit export and retention
- SSO
- Custom policies
What it could be worthEstimate
What a software business making this much sells for today, from real deals. Pick a monthly revenue.
If it makes
Worth about
$190,000
Sale range $132,000 to $263,000
$120,000 a year in revenue
From 87 software businesses listed and sold on Empire Flippers in the last year, revenue vetted by the broker: they sell for about 19× their monthly revenue, most between 13× and 26×.
Measured on profit, the middle deal is 32× monthly net profit: about 2.7 years of profit.
What products in this market make
- Fivetran + dbt Labs combinedEstimate$50,000,000 a month, worth about $950,403,000 at that multiple · sacra.com
- Fivetran (merged with dbt Labs)Estimate$27,083,333 a month, worth about $514,802,000 at that multiple · sacra.com
- dbt LabsEstimate$8,333,333 a month, worth about $158,400,000 at that multiple · sacra.com
An estimate, not an offer or financial advice. A real price depends on profit, growth, churn, how much of the work is the owner's, and the buyer.
How it would growAI proposal
How the product would reach new customers through its own use, and how the first ones would be found. A plan to test, not a result.
- Built-in loop
- Shared outputs. Lineage maps, cost reports and audit exports are shared by link with security, finance and compliance colleagues who do not use the product. Pull request checks also show the product name to reviewers.
- Do they become customers?
- Reviewers and auditors are often not buyers, but some are data leads who could be. Test whether viewers of shared reports or PR checks ask for access and start a trial.
- First customers
- Data platform leads at companies of 100 to 500 people that run dbt on two warehouses and handle health or finance data.
- Where to reach them
- Hacker News and data forums: Share a compile-time lineage demo on an open dbt project, and answer the questions in the original thread.
- dbt and Dagster communities: Offer a free lineage and masking check on public Slack and meetups, then talk to teams who use it.
- Direct outreach on LinkedIn: Message platform leads in regulated industries and offer a 30-minute governance gap review.
- First milestone
- Five paying teams on the Team plan, each running two or more warehouses, within 90 days of the first demo.
Why people already pay for this
- VerifiedHacker News post received 122 points and 48 comments
- Public claimData teams using cloud warehouses lack a DAG control plane that unifies governance, lineage, and cost attribution across multiple warehouse environments.
- Public claimColumn-level lineage is typically computed post-hoc via graph crawls rather than at compile time, limiting precision and real-time feedback.
- Public claimGovernance features (masking, audit trails, classification) are scattered across warehouse-native tools and lack unified CI gates.
Tech stack
Every proposal builds on the same PWA base, installable from the browser on any phone or computer. Dashed boxes are what this app adds.
- AppNext.js and TypeScript
- PWAManifest and service worker
- UITailwind CSS
- Data and authSupabase Postgres
- PaymentsStripe
- EmailResend
- HostingVercel
- This app · SQL parsingMulti-dialect SQL parser libraryColumn lineage at compile time needs parsing for Databricks, Snowflake, BigQuery and DuckDB SQL.
- This app · Warehouse connectorsRead-only metadata and billing connectorsNeeded to read schemas, apply masking policies and pull query cost for each warehouse.
RisksAI inference
- TechnicalAccurate column lineage across four SQL dialects, with macros and templating, is hard. Wrong lineage on governed data would damage trust quickly.
- CommercialTeams may feel dbt, Dagster and warehouse tools already cover enough. The cross-warehouse value must be shown to teams that really run more than one.
- PlatformWarehouse vendors keep extending their own governance and could narrow the gap. API changes may break connectors.
- RegulatoryCustomers may expect compliance claims such as audit readiness. Tracewell can help evidence controls but cannot certify them. Holding warehouse credentials raises security demands.
- DataThe evidence is mostly public claims from one forum thread. Demand among paying teams is unproven and needs interviews.
Everything marked AI proposal or AI inference was written by a model from the evidence on the opportunity page. The proposed name must be checked against trademark registers before it is used.