PROJECT CONTROLS. REBUILT. / IN DEVELOPMENT

Project controls shouldn't run on spreadsheets.

Contarva is the operating system built to replace them.

We're building a connected operating layer for cost, schedule, commitments, progress, change and risk—integrating with systems of record such as SAP, Oracle and Primavera P6, with AI designed to investigate what changed, why, and what needs a decision.

Become a Design PartnerExplore Contarva

Seeking design partners in energy and capital projects. The workflows, integrations and product views below describe the intended product; availability and pilot scope will be agreed with each partner.

01 / THE OPERATING LAYER

Keep your systems of record.
Replace the spreadsheet layer between them.

Excel fills the gaps between finance, scheduling and the decisions a project team has to make. Contarva is being designed to take over that work.

TODAY / DISCONNECTED WORK

One project. Many versions of the truth.

SAP / OraclePrimavera P6Contractor data
Excel extractsCost workbookChange workbookForecast v09_FINAL
Manual reconciliationEmail explanations
PowerPoint → Management

The reporting pack arrives. The assumptions and decisions remain scattered across files and inboxes.

WITH CONTARVA / PROPOSED ARCHITECTURE

One connected project context.

SAP / OraclePrimavera P6Contractor data
CONTARVA PROJECT GRAPH

WBS ↔ cost account ↔ contract ↔ activity
change ↔ quantity ↔ risk ↔ milestone

Controls + ForecastChange + ProgressAI investigation
Decision feed + approval history

A shared structure for the numbers, the evidence behind them and the decisions that follow.

The proposed starting point is controlled file imports, including ERP extracts and P6/XER. Direct connectors, refresh frequency and any write-back are subject to validation and agreed pilot scope. SAP, Oracle and P6 remain the systems of record.

02 / CONTROL → INVESTIGATE → DECIDE

Run the controls process.
Then challenge what it says.

Connected workflows establish the numbers. Evidence-led AI helps the team interrogate them. People make the decisions.

01 / Control

Run the monthly cycle in one place.

The planned controls workspace brings actuals, accruals, commitments, changes, progress and forecast assumptions together. Package owners review exceptions and approve a versioned period close.

Baseline Actual
Illustrative cost variance+8.2%
02 / Investigate

Follow the movement to its cause.

AI-native investigation is being designed to challenge unexplained movements and unsupported assumptions against connected project evidence. Findings should show their sources, uncertainty and the owner who needs to respond.

? Flag raised Package owner Root cause
Illustrative investigationMill lead-time
03 / Decide

Turn an exposure into a decision.

Compare forecast scenarios, record the assumptions behind each option, and bring cost and milestone consequences into management review. The proposed workflow keeps approval and accountability with the project team.

Today
Illustrative contingency scenario62% remaining
03 / THE MONTHLY CONTROLS CYCLE

From period close to management review.

Our initial product focus: replacing the master cost-control and forecasting workbook with a repeatable, auditable workflow.

01 →

Import

Bring in ERP actuals, commitments, P6 schedules and contractor progress through agreed exports.

02 →

Reconcile

Resolve missing codes, period cut-offs, duplicate records and inconsistent mappings before forecasting.

03 →

Forecast

Update estimate to complete (ETC), accruals and package assumptions with named owners.

04 →

Explain

Bridge estimate at completion (EAC) movements to productivity, scope, rates and approved changes.

05 →

Challenge

Test forecast assumptions against observed performance; flag gaps in evidence for human review.

06 →

Review

Give management a concise decision agenda, with owners, approvals and a retained period snapshot.

04 / PROJECT PULSE

A management agenda, grounded in evidence.

An illustrative view of how connected controls and AI investigation could bring emerging exposures into the same conversation as the monthly forecast.

CONTARVA / PROJECT PULSE

North Basin LNG · September 2026

CONCEPT MOCKUP · FICTIONAL DATA
EAC / CAD$2.740B+$38M vs August
RFSU / FORECAST18 May '27+11 days vs August
CHANGE EXPOSURE$86M+$12M vs August
FORECAST CONFIDENCEMODERATEAnalyst rating · evidence gaps

3 items require management attention

Open the evidence behind each illustrative finding.

01 / ASSUMPTION CHALLENGED

Piping forecast assumes an unproven recovery.

Productivity rises from 0.78 to 0.91 in the remaining-work forecast. No approved mitigation supporting that recovery is linked in the sample records.

Potential exposure: $4.1–6.3MOwner: Construction Manager
Inspect evidence and proposed decision
Sources: Progress_W39.csv, package PIP-04; contractor hours, weeks 36–39; Forecast_09.xlsx, assumption A-142.
Basis: earned hours ÷ actual hours = 0.78; forecast assumes 0.91. Exposure range varies recovery timing and remaining hours at the forecast labour rate.
Review: validate quantities and cut-off alignment; request a resourced recovery plan or revise ETC. No forecast change is approved automatically.
02 / POSSIBLE UNREGISTERED CHANGE

Engineering ETC is rising without a linked trend.

Estimate to complete increased across three periods. The sample change register contains no linked trend explaining the additional engineering effort.

Potential exposure: $1.4–2.0MOwner: Engineering Manager
Inspect evidence and proposed decision
Sources: Forecast_07, Forecast_08 and Forecast_09, account ENG-210; Trend_Register_09, revision 3.
Gap: the movement may be reclassification or additional scope; missing linkage alone does not establish a change.
Review: reconcile scope and coding, confirm overlap with the EAC, and assign a trend owner where warranted.
03 / MILESTONE EXPOSURE

Vendor drawings threaten fabrication release.

The illustrative P6 logic links late compressor drawings to fabrication release and the forecast ready-for-start-up (RFSU) milestone.

Potential RFSU impact: 11 daysOwner: Package Manager
Inspect evidence and proposed decision
Sources: Schedule_2026-09.xer, activities CMP-DRW-120 → FAB-210 → RFSU-900; vendor document register, revision 12.
Basis: indicative schedule scenario, subject to logic, calendars, float and recovery review. The 11 days are already reflected in the mockup's RFSU movement.
Review: confirm a vendor recovery date and assess a staged fabrication release with the scheduler.
All figures, files, findings and project details are fictional. Exposure ranges are illustrative, may overlap, and are not additive. Confidence is a qualitative sample rating, not a calibrated probability. This is a product concept, not a live project dashboard.
05 / EVIDENCE BEFORE ASSERTION

Every conclusion. Every assumption.
Every source. Traceable.

This is the design standard for Contarva: a reviewer should be able to follow a finding back through its calculation, assumptions and original records—and see what remains uncertain.

01 / FINDING →

$4.1–6.3M exposure

Illustrative piping recovery challenge, awaiting owner review.

02 / ASSUMPTION →

0.78 → 0.91

Forecast productivity recovery has no linked approved mitigation.

03 / EVIDENCE →

Quantities + hours

Installed quantities, earned hours, contractor hours and remaining work.

04 / SOURCE

Records + versions

Progress W39 · hours W36–39 · Forecast 09 / A-142.

Project Graph + ontology

A proposed shared vocabulary links activities, WBS, cost accounts, contracts, quantities and risks while preserving project-specific definitions.

Controls workflows + history

The intended record includes mappings, period snapshots, assumptions, owner responses and approval history—not only the latest total.

Evidence + domain logic + AI

AI investigation is designed to work against project relationships and controls rules, cite evidence, and expose gaps. Human reviewers retain judgement and approval.

06 / DESIGN PARTNER PROGRAM

Help us rebuild project controls.

We are seeking experienced project organizations willing to map a real monthly controls process and test a focused prototype using an agreed set of historical or current project data.

The project records that make this useful.

Start with a scoped sample. Sanitized historical data is welcome; a complete dataset is not required for an initial conversation.

  • Cost reports
  • Forecast history + assumptions
  • Primavera P6 / XER schedules
  • Commitments + purchase orders
  • Change / trend registers
  • Risk registers
  • Progress / productivity records
  • Monthly management reports

Success also needs context: code mappings, period cut-offs and a controls lead who can explain how the team reaches its forecast.

What we propose to deliver with you.

Each engagement will have a defined scope, milestones and commercial terms.

Founder-led process mapping

Map your monthly close, handoffs, approvals and recurring bottlenecks.

Excel dependency assessment

Identify which workbooks carry critical controls logic and what replacing them requires.

Prototype configuration

Configure a focused prototype around agreed codes, data and workflows.

Historical backtest

Test findings against later outcomes using only records available at each cut-off.

Product influence

Help prioritize the controls workflows, evidence and review experience that matter.

Preferential pilot terms

Agree early-partner commercial terms alongside scope and mutual commitments.

Give Contarva Month 6.

Can the prototype surface an exposure your reporting recognized in Months 7–9? We propose testing that question against historical records, checking false positives as well as useful findings. Earlier detection is a hypothesis to validate, not a promised outcome.

Become a Design PartnerStart with your workflow and the problem you want to solve.
07 / COMPANY

Start with the discipline.
Build around the work.

Spreadsheets exist because project teams need to reconcile imperfect inputs, exercise judgement and explain a forecast. Replacing that layer means understanding the controls process in detail.

Contarva's founder-led approach starts with cost-code mapping, accrual cut-offs, commitment exposure, forecast assumptions and the handoff to management. Design partners will help test that understanding against real project workflows.

Anuj Vij · Founder, Contarva
08 / INVESTOR BRIEF

A focused starting point.
A broader capital-project ambition.

The thesis: the spreadsheet operating layer is where project context and judgement accumulate. Replacing that workflow could create a durable foundation for decision intelligence.

Initial focus

Monthly cost control and forecasting for oil & gas and energy projects. The proposed first proof is a controls cycle run without the master workbook.

Potential expansion

Change, progress, schedule intelligence, risk and portfolio workflows—sequenced around partner evidence and demonstrated value.

Intended foundation

Project Graph, domain ontology, embedded controls workflows and permissioned assumption and decision history. Defensibility remains to be demonstrated.

Request an investor conversation ↗
BUILD THE NEXT CONTROLS CYCLE WITH US

Less time rebuilding the workbook.
More time testing the forecast.

Help define what a project-controls operating system needs to earn your team's trust.

Become a Design Partnerexecutive@vorxi.co