The business case
Data to decision — and what it's worth.
Every number on this page carries its source, its year, and an honesty grade. The math is yours to check.
The talent wall
The people who know how are leaving. The pipeline behind them has collapsed.
US petroleum-engineering BS degrees per year, 2017–2024 — the final point is a projection, not a measurement[4]
$5,475
average cost per hire — all-industry floor (2025)[2]
44 days
median time to fill an open role (2025)[2]
$258,074
US mean total compensation, SPE members (2019)[1]
90%+
of graduates placed before graduation (2023)[4]
And when they leave, the knowledge walks out the door.
No requisition brings back what a thirty-year engineer knew about your fields. There is no citable dollar figure for that loss — which is exactly the problem: it was never written down. SolvxAI’s answer is structural: every study keeps a study memory of what was learned, decided, tried and rejected, and adds to a knowledge map of the wells, formations and reservoirs behind it — both owned by your organization. The next engineer starts from everything the last one learned.
The time drain
Engineering payroll, spent on not-engineering.
~45% of working time on loading and cleansing data[5]
measured across data professionals, 2020
~20% of the workweek searching for internal information[6]
measured across knowledge workers, 2012
Why these two numbers, and not the famous one
Nobody has credibly measured this for petroleum engineers specifically — the claim you may have seen that they spend “50%+ of their time on data” does not survive verification, so we do not use it. These are the closest defensible measurements from adjacent professions, labeled as exactly that.
Whatever the true fraction is on your team, it is the platform’s first target.
SolvxAI takes the not-engineering off the desk: it prepares data on ingest, retrieves what already exists, assembles the working set, and hands your engineers a first draft of the analysis — so their time goes to judgment, not wrangling.
The decision stakes
One decision can outweigh a year of payroll.
Drilled too close
Child wells share the parent’s depleted rock and interfere instead of adding reserves.
Spaced for parity
With enough clearance, each well produces on its own curve.
The evidence is stark. A 2019 SPE study of Midland and Delaware wells (2012–2018) found Permian child wells systematically underproduce their parents, with production parity generally only above ~800 ft of spacing (2019)[7]. Analysts at Tudor, Pickering, Holt & Co. put numbers on the downside the same year: too-tight spacing risks losing 15–20% of ultimately recoverable crude, with child-well recovery running 20–30% below parents across much of the Permian (2019, analyst estimate)[8].
This is why the platform’s job is not saving hours. It is getting the irreversible decisions right — spacing, completion design, the bid — with every discipline at the table.
Your numbers, your math
Build the estimate yourself.
Three pillars, seven dials. The register-anchored benchmarks are marked with their citation; every other input is labeled as your assumption. Nothing is hidden in the formula.
With your inputs, the annual value comes to
$2.2M
per year — capacity, continuity, and decision quality combined
Capacity reclaimed
Team size × loaded compensation × the share of time you believe the platform gives back.
Engineers, geoscientists, and analysts on the asset.
Defaults to the SPE 2019 US mean total compensation — set your own.
The default sits well below the adjacent-profession measurements cited in the footnote.
Loaded comp defaults to the $258,074 US mean total compensation (2019)[1]. Time reclaimed is yours to choose — the adjacent-profession measurements are 20–45% (2020, 2012)[5] [6].
Continuity exposure
What each departure costs you in rehiring and an empty seat — before any knowledge loss.
Expected exits from the technical team in a typical year.
Daily cost of a gap, derived from your comp dial
$707/day × 44 days + $5,475 recruiting (2025)[2] = $37K per departure
Formula: departures × ($5,475 recruiting cost, 2025[2] + 44 vacancy days (median), 2025[2] × your loaded comp ÷ 365). The benchmarks are all-industry US figures — cost-per-hire was $4,129 in FY2015 and $4,683 in 2021[3], so the floor keeps rising. Departures are your assumption, and no knowledge-loss dollar figure is added: none survives verification.
Decision uplift
Production × netback × the uplift you credit to better-informed decisions.
Your asset's total daily oil rate.
Revenue per barrel after operating costs.
Deliberately capped at 3% — set what you believe.
The slider is yours and deliberately capped. For scale of what one spacing decision puts at risk: parent–child parity generally only above ~800 ft (2019)[7], and analysts put 15–20% of recoverable crude at stake (2019, analyst estimate)[8]. We do not claim the platform delivers a specific percentage.
$310K
Capacity reclaimed
$37K
Continuity exposure covered
$1.8M
Decision uplift
Even with the uplift dial at zero, capacity and continuity alone come to $346K a year on your inputs — the decision lever is upside on top of a floor you set yourself.
The evidence register
We grade our own evidence.
Marketing pages usually hide their sources. Ours is the product of a verification pass that discarded every number that failed — including some that would have flattered us. What survived is below, with the grade we gave it.
- 1
SPE — Membership Salary Survey Highlight Report (2019)
Primary — rock solidFigures used: US mean total compensation $258,074; worldwide mean $197,619 (n=4,037, self-reported).
Caveat: 2019 remains the latest edition with a verifiable US mean; the 2021 edition (the latest covered by SPE's journal) reports worldwide base + bonus of $146,861 — a narrower construct, verified July 2026.
- 2
SHRM — 2025 Talent Benchmarking (2025)
Primary — rock solidFigures used: Average non-executive cost-per-hire $5,475; median time-to-fill 44 days (n=2,371, fielded Jan–Mar 2025).
Caveat: All-industry US baseline — a conservative floor for specialized petroleum engineers. Time-to-fill reported as a median in the 2025 methodology.
- 3
SHRM — Benchmarking series (trend anchors) (2015–2021)
Primary — rock solidFigures used: Cost-per-hire $4,129 (FY2015) → $4,683 average (2021, n=472) → $5,475 (2025 [2]): the floor keeps rising.
Caveat: Historical trend anchors only.
- 4
Heinze survey (Texas Tech), reported in SPE's JPT — US petroleum-engineering degree and enrollment series (2016–2024 (reported 2018–2023))
Primary — rock solidFigures used: US petroleum-engineering BS degrees: ~2,615 peak (2017) → 894 (2022) → 655 (2023) → ~500 projected (2024). Undergraduate enrollment 8,712 (fall 2016) → 6,263 (fall 2017). 90%+ of graduates placed before graduation (2023).
Caveat: Realized figures are primary; the 2024 number is a projection and is always labeled as such.
- 5
Anaconda — State of Data Science (2020)
Directional — labeledFigures used: Data professionals spend ~45% of working time on data loading (19%) plus cleansing (26%) (n=1,099).
Caveat: Measures data-science professionals, not petroleum engineers — the closest credible measurement of the wrangling burden.
- 6
McKinsey Global Institute — The Social Economy (2012)
Directional — labeledFigures used: Knowledge workers spend ~20% of the workweek searching for internal information.
Caveat: 2012 estimate for generic knowledge workers; context, not measurement of today's engineers.
- 7
SPE-194310-MS (SPE Hydraulic Fracturing Technology Conference) — Permian parent–child well interference study (2019)
Primary — rock solidFigures used: Permian child wells systematically underproduce parents; production parity generally only above ~800 ft spacing (Midland + Delaware data, 2012–2018).
Caveat: Spacing-conditional; thresholds vary by bench and parent depletion.
- 8
Tudor, Pickering, Holt & Co., reported by Bloomberg / World Oil — Permian child-well recovery analysis (2019)
Directional — labeledFigures used: Too-tight parent–child spacing risks losing 15–20% of ultimately recoverable crude; child recovery 20–30% below parents in much of the Permian.
Caveat: Analyst estimate, not a measured field study.
- 9
Global Energy Talent Index (GETI) — Energy workforce survey (2026)
Directional — labeledFigures used: Average North America oil & gas contractor day rate ~$776/day (9,000+ respondents, 143 countries).
Caveat: Recruiter-run survey; blends all roles and seniority levels.
The savings are the floor. The asset you build is the upside.
The dollars above repeat every year. But every study also deepens your knowledge map — the wells, fields, formations, reservoirs, leases, completions, operators, basins and plays your work touches, with 3D, timeline and sources views — and its study memory: what was learned, decided, tried and rejected. Every report and deck carries the source graph it was built from. It is an asset that appreciates while the software around it depreciates, it survives turnover, and it is built on your data.
See how the knowledge map and study memory workIntelligence per dollar
More intelligence per dollar — and it grows with every job.
The platform evaluates its own work, names what to improve, and carries that into the next job. The same spend buys sharper work over time.
- 01
Every step, judged
Every agent step is judged by fast decision models — System One models — through a gateway with zero data retention.
- 02
Improvement themes
The system surfaces its own improvement themes: where it went wrong, why, and what to change next.
- 03
Recursive self-improvement
Each job tunes the environment the agents work in, so the next job starts from a sharper setup than the last.