When Your Employees Are Your API
Why do aviation distributors and MROs lose time to manual work? Explore connected ERP workflows, real operator quotes, and ways to measure improvement.
TL;DR
When aviation employees become the connection between disconnected systems, manual work consumes their expertise. Real transformation gives distributors and MROs more capacity, with measurable improvements in delivery, turnaround, and margin.
She came in to sell
Your best salesperson has a repair shop waiting for an answer. They need a part by Friday. She knows the buyer, understands the urgency, and knows there's business here. Now watch what happens to her day.
She opens a vendor reply, copies the information, searches old pricing, and asks whether the unit showing in inventory is available. Operations starts looking for the trace. Downstairs, the warehouse guy checks the spreadsheet everybody trusts more than the inventory screen.
Let's call him Tony. You'll recognize him.
By the time the answer comes together, your salesperson has spent her morning assembling information. The buyer is still waiting. Somewhere else, a technician is waiting too. To follow the problem through, picture a distributor supplying a repair facility. That connecting transaction is illustrative; the operator accounts here come from separate, anonymized conversations.
One brokerage leader reported 500 quotes a day across five salespeople. Later, describing the pressure on his team, he said:
"We can't expand, and I don't want to throw more bodies at it"
His people were working. Hard. The process was consuming what they could accomplish.
We've got to stop treating that as proof of a dedicated team and move on. Dedication deserves a better operating environment. When somebody has to retrieve, translate, and pass information between disconnected systems, you've made that person the connection. An API lets systems request information or actions from each other. In your business, the request goes to a person who's supposed to be doing something else.
Your employees are your API. Every transaction takes another piece of their day.
The invoice you never see
Then the owner looks at a software proposal and says the company is too small.
Too small based on what? The number of people on payroll tells you very little about the transactions, documents, and exceptions those people handle. A small team can carry a complicated operation. It can also bury an astonishing amount of labor inside routines nobody has priced.
For illustration, take five people losing three hours apiece each day: an hour re-entering information, another searching and reconciling it, another chasing the next step. Across 240 working days, at a loaded labor cost of $43.60 an hour (Company A), that's $156,960 annually assigned to the workaround. Put an annual system cost beside it. These are random figures, they’re real returns.
Run your own numbers. Include implementation, training, and ongoing costs. Some work will remain. But count what you're already funding before deciding you can't afford to change.
The missed sales conversation sits outside that labor calculation. So does the order lost because you answered late. Verify those consequences separately. Adding guessed revenue to payroll doesn't create a business case.
There's evidence that assistance can increase capacity: a study of customer-support agents found a 15% average increase in issues resolved per hour with AI assistance. Its relevance here is the mechanism, helping people complete more work, rather than a percentage to paste into an aviation proposal. Generative AI at Work​
And please, don't blame Tony. He built that spreadsheet because something needed protecting. Find out what. Then put that knowledge into a process the warehouse can maintain and other authorized people can use. Otherwise, you'll buy new software and still call Tony every time you need the truth.
Carry the deal beyond the salesperson
Go back to that Friday requirement. Give your salesperson a workflow that brings the request, eligible stock, vendor offers, pricing history, and documentation context together. In ERP.Aero, supported RFQ intake can structure incoming demand, and the quoting workflow puts those choices in front of the person making the decision.
She's checking the condition, the evidence, the supplier's commitment. She's judging whether the deal makes sense. Configured pricing and approval rules handle the routine boundaries. Her experience goes into the decision instead of recreating the file around it.
When the customer accepts, carry the requirements into the linked sales order, purchasing, receiving, and fulfillment records. Tony can see what should arrive and where it's going. The inspector can see what was required. A box reaching the dock remains distinguishable from material accepted for use.
Now change the supplier's delivery date. Can purchasing establish which customer promise is affected? Can sales evaluate another source with the cost and margin in view? That moment tells you more than a polished dashboard ever will.
Suppose the cheapest offer is old and the alternative can meet Friday. Your salesperson needs to see the difference before committing. A lower buy price has little value if nobody has confirmed availability. That's where her judgment belongs, weighing a real option with the customer, while the supporting records stay connected.
Another distributor participant challenged us over customers having to copy tracking information into a carrier website:
"It's just work for a very high value client that I don't think is necessary."
Fair criticism. The customer experiences the work we leave them, including the small inconveniences. Follow that standard through delivery, the invoice, and any remaining return, credit, or core obligation.
Here's the proof I'd want from this distributor: less time to a valid quote, fewer eligible requests left unanswered, fewer fulfillment corrections, and dependable delivery at understood margins. Track human handling across departments. If sales saves time while receiving reconstructs the same information, you've moved the burden. In a recent conversation with Oleg, I put the test simply: “What work disappears after the project is finished?” Carry that question through the transaction, including the work waiting for the next department.
And the consequence of that burden is now waiting at the other end of this order, on the repair bench.
The part has a job waiting for it
Walk back to when the customer's unit first reached the repair shop. One operator described the beginning plainly:
"We don't know when parts are coming in. They might send us a notification, might not, but a part, hits the door."
Out comes the component. Receiving checks the part and serial number against the customer repair order. Somebody establishes what work was requested, where the unit belongs, and whether the documents match. That uncertainty needs handling before the shop can do useful work.
In another repair discussion, a mechanic I know described printing work orders, technicians completing them by hand, and information subsequently being entered digitally. Look at the people involved. A technician records the work. Somebody else reads it and types it again. Quality then depends on those records when it's time to inspect and release.
Every additional translation gives information another chance to arrive late or wrong.
A better route starts with verified identity and carries it into the work order. Incoming documents stay attached to the job. Inspection findings establish the actual scope, material, and labor requirements. Customer authorization remains connected to what the shop is permitted to perform.
ERP.Aero's repair workflows include checklists, role sign-offs, labor and material records, and closeout reporting. Put them to work where the task happens. If the technician still needs a second person to reconstruct the day's work, the implementation hasn't removed that dependency.
Measure receipt-to-job readiness and the effort needed to create a complete record. Include mismatches and missing documents. Getting a clean box through quickly is useful; resolving a difficult receipt without losing control is where the process proves itself.
Then teardown finds the missing material. Now the shop needs the distributor's Friday promise to become something it can actually plan around.
Keep the repair connected to the buying
The repair conversations made this dependency impossible to miss. Operators described printing availability lists, identifying shortages, gathering vendor prices, obtaining approval, purchasing, and issuing material. One colleague stopped the purchasing discussion at the prerequisite:
"before we even get that, we need to price. So we have 40 parts in there. We gotta get that priced."
Of course. The shop needs a defensible customer estimate before it can proceed with authorized buying. Having a purchase-order function doesn't settle that chain of decisions.
The requirement was to source with the work order's context intact. That applied to our demonstration too: some sourcing moved into ERP.Aero's connected supply-chain workflow. We had to account for the movement and what the operator needed, rather than pretending everything happened on one screen.
Link the required material to allocation, supplier offers, purchasing, receipt, acceptance, and issue. Production control needs to establish what's actually missing and which promise depends on it. As one operator put it:
"one little O ring or a screw is holding up a work order, and we don't want to lose sight of that"
That little item can hold the whole job. Track the waiting it causes alongside time spent on the bench. Then follow completed work through inspection, authorized release, shipping, and invoice readiness. Evidence should develop with the repair so closeout doesn't become another search exercise.
When the promised part reaches the dock, the job still needs an answer: has it passed inspection and become available to this work order? If it's on hold, production control needs the reason and the person resolving it. Marking a purchase order delivered won't get the assembly back on the bench. The physical event has to reach the people planning the repair.
Technology can make this worse if it produces confident answers without the underlying checks. A GPT-4 experiment with BCG consultants found roughly 19 percentage points lower correctness on a task outside the tested AI's capabilities. That belongs in the evaluation: challenge the answer and verify the action. Navigating the Jagged Technological Frontier​
AOG urgency doesn't authorize missing inspections. A generated record doesn't prove the work happened. Show the evidence, including when something fails.
Give people their jobs back
That standard belongs on both sides of the implementation table. Owners have to make time for their operators to help design and test the work. Familiar routines need examination. Providers, including us, have to show what happens when the easy demonstration ends.
Bring the incomplete vendor reply. Change the promised date. Send the wrong serial number. Correct a transaction after the next department has used it. Make us explain what's configured, what requires development, and how the person responsible gets the work moving again. Don't accept a demonstration that quietly relies on somebody outside the system to finish it.
Then compare similar work before and after. For the distributor, follow the quoted opportunity through profitable delivery and remaining obligations. For the shop, follow the serialized unit through turnaround, rework, complete release records, actual job cost, and billing. Include training, maintenance, corrections, and the experienced person everyone interrupts for help.
Returned time creates capacity. Lower spending establishes cash savings. Additional orders establish commercial improvement. Keep those results distinct so everyone can see what actually changed.
Now return to Monday morning. Your salesperson has the context to make the call and work the opportunity. Tony spends his time improving location accuracy and controlling movements, with records other people can trust. Operations can establish what's waiting and act on it. The technician can concentrate on the authorized repair while its required evidence stays connected.
That's the improvement to demand from an aviation system, including ERP.Aero. Walk the transaction. Make the difficult case work. Ask the people doing it whether their day has changed.
You hired a salesperson because she can win business. Give her back the part of the day where that happens.
Pick one transaction your team keeps rescuing. Follow the copying, searching, and chasing through every department. Price that effort. Then make the improved workflow prove itself with the people who’ll use it.
Your best salesperson should have time to sell. Give her that opportunity.
About me and ERP.Aero
I didn’t start in aerospace and defense, but that’s where my career brought me. After nearly two decades working around aviation parts, sales, sourcing, distribution, repairs, and the systems behind them, I focus on removing unnecessary work so good people can use their judgment.
​ERP.Aero is an aviation-native operational platform connecting quoting, sourcing, inventory, certification, repair, fulfillment, finance, and embedded intelligence. iRFQ helps bring inbound demand into the workflow. ELIA is designed to apply intelligence within the transaction, where the context and decisions belong.