Money leaks where the field and the office disagree.
Your ERP shows the invoice. Your trackers show the truck moved. Neither shows who ran the machine, what it produced, what it burned, or why it stopped. That gap is where your margin goes.
Darikoda is the operating record built around how your operation actually runs, so it gets used, the leak closes, and you scale on numbers you can defend. Captured at source.
Keep the leakage map regardless of next steps.
184 / 184
CFuel dispenses captured at source
Every litre logged at the machine with photo, GPS and shift. The fuel number can be trusted.
Fuel recovered
AGHS 18,420
Fuel variance exposure
BGHS 61,500
EX-22 coolant
C103 C
Health alerts open
C3
Fleet utilisation
C72%
Parts variance
B5%
EX-22 excavator
Health alertCCoolant climbed from 90 to 103 C across the shift. Three health alerts open on the asset.
Darikoda is operational intelligence software for Ghanaian fleet-heavy operators. It is the operating record built around how your operation actually runs: a time-stamped record of fuel, machine hours, downtime, approvals and certificate evidence. So contractors keep operations predictable and defend margin instead of reconstructing events from WhatsApp, trackers and spreadsheets.
In 30 seconds
GHS 18,420
What you get: predictable, accurate operations and the control to scale without the leak. Recovered billable hours, faster cert prep, less fuel variance, fewer disputed deductions.
4 sectors
Who it is for: Ghana fleet-heavy operators in mining, construction, civils and plant hire.
30 min
First step: a free 30-minute audit that maps where your operation is leaking. One page. You keep it either way.
From the field
I have sat inside these operations. Here is what I almost always find.
The leak is rarely where the spreadsheet says. It hides in the gap between what happens on site and what reaches the office. Across mines, quarries, road and construction sites, the pattern repeats.
Your biggest costs are the weakest evidenced.
Fuel that leaves the bowser but cannot be tied to the machine that burned it. Parts and tyres bought with no proof of what they returned. A cost per tonne or per hour nobody fully trusts.
The month-end count decides what you get paid.
The work was done. The other side's record says less. Hours, litres and quantities walk at every reconciliation, because the dispute is settled by whoever holds the structured count, not by who did the work.
Nobody can answer the question that runs the business.
Which machine earns and which one bleeds. What a project actually cost to run. Most operators cannot say, because the numbers to answer it were never put together.
The whole operation runs through one person's phone.
More approval requests than one person can track, most of them on WhatsApp. A cheap part waits days for sign-off while the machine it belongs to stands idle.
The machine comes back worn, and no one owns the wear.
When the operator does not own the asset, the wear sits on no one's record. It shows up later as higher maintenance and machines retired early, quietly capping a margin nobody can explain.
The averages hide the bottleneck.
The fleet looks busy on average. Split it by machine class and you find one type sitting idle while a single bottleneck holds up everything behind it.
Your best controls already exist, in someone's head.
A fuel log a supervisor keeps by hand. A materials count someone invented. The discipline that lives in one experienced head walks out the door when they do, and takes your only control on the biggest cost lines with it.
Most operations software is pretending to work.
Not because it is bad software. Because your ERP, your trackers and your dashboards cannot see what happened in the field: who ran the machine, what it produced, what it burned. The licence renews while the real record stays in heads, on paper, and in WhatsApp. That gap is not a software failure. It is the layer no one built, and it is the layer Darikoda is.
Captured where the work happens.
The day's reality, logged at the moment it happens, not reconstructed from memory at month-end.
One operating record, seen through your lens
This works across heavy-fleet operations. We build it through your reality.
Pick your operation, then your role. These are the questions a 30-minute audit starts to answer for an operation like yours, with your own numbers.
Start by picking your operation above.
Each lens shows what the field says an operation like yours is most likely leaking, and routes you to the page built for your role.
184 / 184
Fuel dispenses captured at source
Fuel recovered
AGHS 18,420
Fuel variance exposure
BGHS 61,500
EX-22 coolant
C103 C
EX-22 excavator
Health alertCoolant climbed from 90 to 103 C across the shift. Three health alerts open on the asset.
In mining, your margin is in the hours, the fuel and the tonnes you can prove. Most of the leak is in what you cannot.
- Can you prove every billable hour the mine's QS challenges, to the minute?
- When fuel goes missing, can you name the machine that burned it?
- Do you know your true cost per tonne, by truck and by haul road?
USD 285,000
Money at risk from time, Parkview Phase 2
Spend ratio, EAC
C108%
Forecast profit
C-USD 115,000
Margin erosion
B-USD 2.87M
Parkview Phase 2
SPI 0.71SPI 0.71, 19 days late. USD 132,000 exposure on the flagged phase, driving the loss.
Across a portfolio, the project that is losing money hides inside the average.
- Which of your projects is quietly losing margin right now?
- Does materials variance reach you before the crew moves on, or after?
- Could you defend a subcontractor deduction with evidence today, not paper?
GHS 82.40
Cost per usable tonne, TR-07
Fleet idle
C43%
Addressable idle hrs
B1,240
Idle exposure
BGHS 310,000
CR-02 crusher
Health 58Health score 58, an open alert. Waiting on parts for about 260 days while output runs down.
Your cert is your cash. The leak is in what you cannot defend when it gets cut.
- When the consultant cuts a quantity, what can you push back with?
- How much faster would your IPC pack be if the evidence was already structured?
- Where in the cycle is your cash actually getting stuck?
USD 64.20
Best lifetime cost per hour, OEM A
OEM B per hour
CUSD 88.50
Addressable idle
C19%
Owning cheaper
C4 of 5
OEM A grader
Best valueOwning works out cheaper than hiring on four of five assets on the current fleet.
Your machine earns on someone else's site, run by an operator you do not employ. The biggest leak is the wear you never see coming.
- When a machine comes back worn, can you say what caused it, and recover it from the sub?
- How many rental days a year vanish to breakdowns that did not have to happen?
- Could you defend the disputed hours and fuel today, to the litre?
The real reason this matters
Make running your own operation profitable again.
Without the right systems, running your own fleet can quietly cost more than handing the work to subcontractors. That is the leak at its largest: the point where an operator gives up margin just to escape the management burden. Darikoda turns weak field evidence into protected margin, so self-operating stays the profitable choice as you scale.
Margin per fleet, hire vs own
(Illustrative model, your number from the audit.)
Your operation. Your margin.
Self-operating stays the profitable choice when the field record is strong enough to defend the margin.
What it returns.
What the operating record shows you, live:
1 day
Cert packs that come together in a day, because the field evidence is already structured.
GHS 18,420
AFuel variance the day it appears, at source, instead of absorbed into operational noise.
1,240 hrs
BWhich part of downtime is really the largest, broken down instead of logged as one number.
4 of 5
CHire-versus-own and renewal decisions made on per-asset evidence, not memory.
184 / 184
CDisputed deductions defended with evidence captured at the event, not paper timesheets reconstructed at month-end.
Your specific number depends on your fleet, your contracts, and your current baseline. The Operational Audit produces it.
What the platform actually does
Follow one fuel event through the record.
The same flow carries machine hours, downtime, materials and approvals. Fuel is just the clearest example.
01 · Captured at source
The event is logged where it happens.
A fuel event at the bowser, logged the moment it happens: which machine, how much, who was at the controls, with proof attached. Not reconstructed from dockets at month-end.
PINGPSPhotoTime-stamp02 · Held offline
The record holds without signal.
The event is saved on the spot and reaches the office when the signal returns. The operator never waits for 4G.
SavedQueuedSynced03 · Variance flagged
The gap surfaces the day it opens.
Litres dispensed against hours run, per machine. The variance that used to hide for three weeks is flagged while you can still act on it.
Litres vs hoursFlagged04 · Exported as evidence
Month-end opens with the pack ready.
Time-stamped, attributed, immutable. The same record bundles into the cert or deduction defence the month-end argument is settled with.
Time-stampedAttributedImmutable
A schematic, not a screenshot. What the record does matters more than what the screens look like.
Built so the record survives the site.
Offline-first, attributed, immutable. The record holds when the operator moves on and the month-end argument starts.
What that means in practice.
Six commitments, built into how every field event is written, so operations stay predictable and you keep control as you scale.
Every field event is saved on the spot and syncs when the signal returns. No spinning wheel of death, no falling back to paper when the signal drops. Your operators never wait for 4G.
You can always see whether a field entry has reached the office yet. No more "did it go through?"
Every action is attributed to a person, role, device, and time. No silent edits to history.
Shared tablets use PIN-level worker attribution. No one is logged in as someone else.
Failed syncs create visible issues, not silent gaps.
Finance and operations see the same record. Your existing accounting or ERP gets fed the per-machine truth it cannot produce on its own.
For engineers: how it's built →
Why rollouts fail, and why this one is built not to
Most operations software is bought, then never used.
The hard part was never the features. It is getting the crew to pick the system up and keep using it after launch week. You have probably bought software before that the field quietly abandoned. That is the risk we engineer against first.
70%
of digital transformations fall short of their goals, on adoption not features.
McKinsey
Half
of software licences go unused, over a fifth of them pure shelfware.
Vertice
42%
makes construction the second most technology-resistant industry there is.
Yooz 2025
23%
of frontline workers feel they have the tools to do the job. The rest run on group chats and memory.
Frontline research
The pattern is consistent. The failure is rarely the software. It is the rollout, the people, and data that was never clean to start with. That is the part almost nobody resources, so we treat adoption as its own discipline, with the same energy as the system itself.
Built to meet the operation where it is.
The best Western systems land on a Ghana site and simply do not work. They assume reliable power, a desk, clean master data, and office literacy. Ours is built for the opposite.
Built for dumsor. A pit with no signal for a full shift keeps running and syncs clean when the signal returns. No falling back to paper.
Two-tap capture at source on shared, gloved-hand devices. Every event attributed to a person, with no one logged in as someone else.
Messy data is the designed starting point, not a blocker. You do not tidy years of records before the field can begin.
The system surfaces the gap without blaming the person who logged it, so the crew keeps using it instead of hiding from it.
Adoption is a paid step, not a hope.
We productize the rollout so it is owned, not left to chance.
PAID BUILD AND ACTIVATION
We configure the operating record to how your operation actually runs and field-test it on real devices, timed to your fleet, not a fixed clock.
READINESS-GATED
The pilot clock does not start until a jointly signed readiness sign-off. You move forward as it proves out, not before.
AUGMENTS, NOT REPLACES
It feeds your existing ERP the per-machine truth it cannot produce on its own. No rip and replace, no integration project.
A rollout backed by real change management is seven times more likely to succeed, 88 percent against 13 percent.
Prosci
Everyone funds the software. Almost nobody funds the adoption. That is why it fails, and that is exactly what we specialise in. And we are paid on the value the system provably delivers, in your own numbers. Skin in the game, by design.
Why operators trust the record
Built in Ghana by an engineer who lived inside the operation.
Engineer-built, not agency-built.
UCL MSc Mechanical Engineering, then GE precision turbines and Caterpillar/Unatrac across Ghana and Nigeria heavy fleet operations. The operating record was designed by someone who stood at the bowser. Problem first, then the system that simplifies it.
Ghana operating reality first.
L.I. 2431 procurement discipline, FIDIC red book joint measurement, the sliding-scale royalty regime, parts lead times from Tema, Starlink on patchy 4G. Built for this reality, not localised after the fact.
Honest about the stage.
Founding-customer engagements are active across Ghana fleet operations. We would rather hand you a leakage map than a logo wall.
Stress-tested to hold up.
The controls that protect your money and your record were checked hard, on purpose, by an independent pass looking for the ways they could fail. What survives that is a record you can put in front of an auditor, a client, or a court with confidence.
Darikoda is built by Darikoda Systems Ltd, registered in Ghana, with Ilori Streamline Ltd (UK Company No. 16378536) behind it.
It starts with a 30-minute audit.
One page that maps where your operation is leaking. You keep it regardless of next steps.
What a leakage map looks like
One page. Your leaks, named and ranked.
Example format only. Your audit uses your own numbers.
Fuel
- Evidence gap
- Litres not tied to a machine or operator
- Commercial risk
- Variance absorbed at month-end
- First fix
- Dispense record with operator PIN and machine
Hours
- Evidence gap
- Tracker and timesheet disagree
- Commercial risk
- Billable hours disputed
- First fix
- Shift-level machine-hour record, attributed
Downtime
- Evidence gap
- Cause recorded after the fact
- Commercial risk
- Idle cost buried in a single breakdown line
- First fix
- Downtime reason captured at the event
Cert evidence
- Evidence gap
- Chainage proof reconstructed late
- Commercial risk
- IPC cut or delayed
- First fix
- Section-level evidence pack
The 30-minute audit produces this for your operation, ranked by likely commercial impact. You keep it regardless of next steps.
The path from leak to defendable record.
Start with the audit. The first 30 minutes is free.
Free audit
A free 30-minute map of where your operation is leaking. One page. You keep it regardless of next steps.
Full audit
A paid, in-depth audit that quantifies the leak across your fleet, contracts and sites, and sizes the opportunity.
Build & activate
We configure the operating record to how your operation runs and field-test it on real devices. Timed to your fleet, not a fixed clock.
Pilot, then rollout
A pilot proves the value in your live contracts. Then an ongoing subscription.
Founding-customer engagements active across Ghana fleet operations. First Build & Activation phase now underway. First published case results targeted Q3 2026.
Ready to see your own leakage map?
30 minutes. WhatsApp. You keep the map either way.
What's already real.
No logo wall and no inflated case studies. The honest state of where Darikoda is today, and what you would be joining.
Early build is underway with founding customers in Ghana.
Build and activation now. Published results when they are ready.
We are in early build and activation with founding customers in Ghana. We configure and prove the operating record on your operation before any ongoing commitment begins.
Published results will come from these first engagements, with their permission, when they are ready.
If yours is next, you get founding-customer pricing and my direct line for two years.
Talk to me
The fastest way is WhatsApp.
+44 7984 845440
theo@iloristreamline.com
Typical reply within the hour during UK and Ghana business hours.
Patterns described here are drawn from extensive field audits and industry research across Ghana's mining, construction, roadworks, and quarry sectors. No specific operator is named or identifiable.
Frequently asked questions.
What exactly does Darikoda do?
Darikoda is the operating record between what your fleet did and what your office can defend at month-end. Fuel events, hour-meter attribution, mechanical availability, materials variance, approvals, certificate evidence. Captured at source.
Does it work in low-connectivity environments?
Yes. Every field write saves locally first and syncs when signal returns. Operators do not wait for 4G. Sites with Starlink plus patchy 3G/4G are well within the operating envelope.
Is this just for mining?
No. Mining is one of five verticals. Darikoda also serves construction (developers, main contractors, subcontractors), civils and roadworks (government contractors under FIDIC), and plant and equipment hire firms. The operating record architecture is the same across all of them.
Why start with an audit?
Because most operators already know money is walking out. The audit is 30 minutes that produce a one-page leakage map. You keep it regardless of next steps. It pre-qualifies whether Darikoda is the right next move or whether something simpler fixes the issue first.
How fast can someone speak to us?
Usually within the hour during UK and Ghana business hours. WhatsApp is the fastest channel. Email works for asynchronous follow-up.
Operating Notes
Latest from the field analysis.
Construction · Concrete24 July 2026The concrete argument is over the moment the pour sets. The record has to exist before then.
Read the full note- Cross-vertical · Operational records23 July 2026
Site diaries: the operational record that decides whether your project can be defended.
Read the full note 
Cross-vertical · Utilisation23 July 2026Equipment idle time: the most expensive hours on site are the ones nobody records.
Read the full note