Your filings are buried in
other people’s PDFs.

We build systems that read the statements, contracts, and filings your business runs on, turn them into records you can file and audit, and flag anything the machine isn’t sure about before it reaches a regulator.

Provider statement, as received

PURCHASER STMT 09/2026   PROP 4412-A

GROSS MCF ........ 14,207.40

BTU FACTOR ........ 1.0413

SETTLE PRICE ...... 2.8140 /MMBTU

TRANS DEDUCT ...... (1,842.06)

OWNER INT ........ .0781250

Extracted and checked

Sales volume
14,207.40 Mcf
BTU adjustment
1.0413
Settlement price
$2.8140
Transportation allowance
$1,842.06
Owner interest
0.0781250
Royalty due
$3,108.49
Matched to source Held for review — allowance exceeds the prior three months

Built for firms too small to staff a data team

A large prooducer has people inhouse for this. A ten person operator has the same reporting obligation and no one to hand it to. That gap is the whole business.

Document extraction Read recurring statements, contracts, and filings in the formats you already receive.
Validation Check each value against its source and stop uncertain records for review.
Workflow integration Run the system where your data is allowed to live and fit the process you already use.
Audit trails Keep every filed value traceable to the page and document it came from.

How an engagement runs

We look at what you actually do today

A few real documents and an hour with whoever currently handles them. Most of the value is in seeing where the time and the errors go.

We build the extraction on your documents

Not a generic tool. A model tuned to the formats your providers and counterparties actually send, running where your data is allowed to live.

We build the part that catches mistakes

Every value is scored against its source and checked for consistency. Anything uncertain stops and waits for a person. This is the part that matters.

You run it, we stay reachable

Handover with documentation, then ongoing support if formats change or volumes grow.

A recent engagement

An Oklahoma operator with eleven wells on a western Oklahoma asset was preparing federal royalty reporting by hand, working from monthly statements sent by two midstream providers in two different formats.

We built the pipeline that reads both formats, extracts the volumes, prices, and deductions, checks each value against the statement it came from, and holds anything inconsistent for review before submission. The model runs inside the client’s own environment, so no statement leaves their control.

1 week → 1.5 days Monthly preparation time
80% Reduction in turnaround
2 formats Reconciled to one record
Every value Traceable to its source page

What it costs

Fixed scope, quoted before we start. No hourly billing and no open-ended engagements.

Workflow assessment

From $7,500

Two weeks

A written answer to whether this is worth automating, before you commit to building anything.

  • Review of your current process
  • What can be automated and what can’t
  • Estimated time and error reduction
  • Fixed-price proposal if you go ahead

Pipeline build

From $30,000

Six to ten weeks

A working system for one document type, running in your environment, with the review controls built in.

  • Extraction tuned to your formats
  • Validation and confidence scoring
  • Review queue for flagged records
  • Audit trail and documentation
  • Training for whoever runs it

Ongoing support

From $2,500

Per month

For when a provider changes a format or your volumes grow. Optional, and month to month.

  • Format changes handled
  • Accuracy monitoring
  • Priority response
  • Quarterly review of what’s flagging

Every engagement starts with a conversation about whether it’s worth doing. If the answer is no, we’ll tell you that before you spend anything.

Who you’d be working with

Pivot Core Group is led by Michele Tchindje. Before starting the practice she spent four years inside a Fortune 100 company doing analytics and AI implementation, building reporting used across all fifty of its divisions and working in aircraft maintenance, cargo, and enterprise planning data.

She works on problems where a wrong number has consequences. That tends to mean regulated reporting, contracts, and anything that eventually gets audited.

Based in
Atlanta, Georgia. Working with clients in Oklahoma and Texas.
Focus
Analytics and AI implementation in regulated and document-heavy work.
Background
Industrial and Systems Engineering, University of Oklahoma.
Sectors
Energy · aviation · manufacturing
Also
Former International Chair, National Society of Black Engineers

Tell us what you’re doing by hand.

If it involves reading the same kind of document every month, there’s usually something here. A first conversation costs nothing and takes about thirty minutes.