Runs on a schedule
Nightly, hourly, or more often — whatever the report needs and the source can handle.
Data warehousing + ETL
Get your ERP data into a warehouse you can actually report on.
I build the pipelines that pull from your ERP, payroll, CRM and the spreadsheets around them on a schedule, land the data cleanly, keep its history, and model it for reporting — in a warehouse I stand up in your cloud, or the one you already have. A reporting solution can go on top, or the warehouse is handed to your own team and BI tools.
- 2–4 weeks, ahead of a report or on its own
- Vista, CMiC, JDE, Epicor and more
- In your cloud, owned by you
Last night’s loads
Ran 02:00 · next run 02:00 tomorrow
| Source | Rows | Status |
|---|---|---|
| ERP · job cost detail | 48,210 | Loaded |
| ERP · commitments & change orders | 3,904 | Loaded |
| Payroll · hours & labor cost | 12,677 | Loaded |
| CRM · accounts & opportunities | 1,138 | Loaded |
| Estimating workbook · SFTP drop | — | Late · alert sent |
Row counts, duplicates and job-cost totals reconciled to the source: passed
Sound familiar?
The month-end your data costs you today.
- Reports run straight against the ERP and slow it down
- Exports stitched together in Excel every month-end
- Job cost in the ERP and hours in payroll never agree
- No history — once the period closes, last month’s numbers are gone
What I build
A scheduled, monitored path from your source systems to clean reporting tables.
1 · Extract
From your systems
Direct database access, vendor APIs, or scheduled exports — whichever your ERP or CRM allows. Incremental where the source supports it, so loads stay fast.
2 · Land
Raw, as-is, with history
Every load is kept, so you can always answer “what did this look like last quarter?” — something the ERP itself usually can’t.
3 · Transform
Cleaned, matched and connected
One job, one cost code, one employee, one calendar — across the ERP, payroll and CRM. Business rules live here, documented and versioned, not in a macro.
4 · Serve
Reporting tables
Wide, fast, understandable tables built for reports. Ready for a reporting solution from me, or for Power BI, Excel and anything else that speaks SQL.
Monitored, with alerts
You find out a load failed from an alert — not from a manager asking why the numbers look wrong.
Data quality checks
Row counts, nulls, duplicates, and totals reconciled against the source after every load.
Sources
The systems I pull from.
If a system has a database, an API or an export, I can get data out of it. The ones below I already have.
ERP
Job cost & accounting
Jobs, cost codes, commitments, GL, AP/AR, equipment
ERP
Manufacturing & inventory
Work orders, production, purchasing, stock
Payroll & HR
Hours and labor cost
Timecards, crews, rates, burden
CRM
Customers & pipeline
Accounts, opportunities, bids, support cases
Finance
Expenses & budgets
Expense reports, budgets, actuals
Databases
SQL Server · Postgres · MySQL · Oracle
Any application or custom system database
Files
CSV · Excel · SFTP drops
Vendor feeds, estimating workbooks, legacy exports
APIs
REST & SaaS
The cloud tools that only talk over an API
Experience integrating data from
ERP
- Vista Viewpoint
- CMiC
- JDE
- Epicor
- Coins
- NetSuite
- Dynamics
- FACS
CRM / Other
- Zoho
- FreshDesk
- Compass
- Lasso
- ADP
- Concur
Don’t see yours? If it has a database, an API or an export, I can pull from it.
Don’t have a warehouse?
I’ll stand one up — right-sized, in your cloud, owned by you.
A data warehouse for a construction or manufacturing business doesn’t need to be a platform project. I set up the database, its structure, the data feeds and who can access what, then start landing data — the whole build is 2–4 weeks. Already have a warehouse? The pipelines load into that instead.
- Your choice of platform. SQL Server, Postgres, Snowflake or BigQuery — matched to your team’s skills and budget.
- Raw, clean and reporting layers. A simple, standard layout your own people can understand and extend later.
- Security and access from day one. Service accounts for loads, read-only roles for reporting, no shared passwords.
- Documented and handed over. Every table, every pipeline, every rule — written down. No lock-in.
The warehouse itself typically runs $30–$250 a month, billed by your cloud provider in your account — no platform fee on top.
Warehouse layout
raw
An exact copy of each source, every load kept
ERP · payroll · CRM · files
clean
Cleaned up, duplicates removed, matched across systems
jobs · cost codes · employees · dates
reporting
Wide tables built for the reports you asked for
job cost · WIP · production
Ways to start
Start where you are.
On its own
A warehouse and the pipelines that fill it
For a team with its own BI tools or its own analysts. I build the warehouse and the ETL, document it, hand it over, and stop there — no reporting from me at all.
- Price
- From $10K
- Timeline
- 2–4 weeks
Before a reporting solution
Warehouse and ETL, then the report on top
The data isn’t in a warehouse yet and the report is the goal. The warehouse and pipelines come first; the reporting solution goes on top, so the warehouse proves its value in the same engagement.
- Price
- From $10K + reporting from $15K
- Timeline
- 2–4 weeks, then a first look at the report 1–2 weeks after
Into the warehouse you have
Add a source to your existing warehouse
You have a warehouse; you need the payroll system, the CRM, or the new division’s ERP in it too. I build the pipeline, the monitoring and the reporting tables for that source, and hand them over.
- Price
- From $5K
- Timeline
- Per source, depending on how it exposes its data
Warehouse and ETL work is fixed-scope: anything beyond it is a new statement of work. Ongoing support is its own agreement if you want it — or your team takes it over, since you own it. See all pricing
Questions about pipelines
Will this slow down my ERP?
It shouldn’t. Loads are incremental where the source allows it and scheduled for quiet hours, and once the data is in the warehouse, the reports run against the warehouse — not the ERP.
What if my ERP vendor doesn’t allow database access?
Then I use what it does allow: the vendor’s API, or scheduled exports dropped somewhere the pipeline can pick them up. Which of those your system offers is one of the first things I confirm on the scoping call, and it is priced on that.
Do I need a data engineer to maintain it?
Not day to day. The loads run on a schedule, alert when one fails, and every table, pipeline and rule is documented and handed over. When a source changes — an ERP upgrade, a new company in the ledger — the pipeline has to change with it: your own team can do that from the documentation, or bring me back under a new or extended contract.
Can other tools use the warehouse too?
Yes. It is a standard database in your cloud: Power BI, Excel, Tableau and anything with a SQL connector can read the reporting tables, through read-only roles so nobody reports against the raw loads.
Can you build just the data warehouse and ETL?
Yes. It is its own offering, not the first step of a bigger one. I build the warehouse and the pipelines that fill it, hand them over with the documentation, and stop there.
What does it cost to run?
Aside from the data warehouse, hosting averages under $50 a month. The warehouse itself is somewhere from $30 to $250 a month, depending on the size and scope of the data. Both are billed by your cloud provider, in your account — there is no platform fee and no per-seat license on top.