From the founder · Agency operations
Last updated: August 19, 2026 · Written by Ethan Kramer, founder of Frederick Interpreting Agency, a Deaf-owned interpreting agency serving Maryland, DC, and Virginia. This is the story of why we built our own scheduling and billing platform, and what we learned about what agency software has to do.
Every interpreting agency runs on software, whether it admits it or not. For most of us that starts as a spreadsheet and ends somewhere we never planned. Frederick Interpreting Agency went through every stage, and in 2025 we did the thing agency owners joke about and built our own. Here’s what the three stages look like from the inside, what agency software actually has to do, and what changed for our coordinators, interpreters, and clients when we finally had the right tool under the work.
Dates are approximate; the order is the same at nearly every agency we’ve talked to.
The short version
- There are three software stages, and each one breaks in a predictable way.
- Interpreting has its own shape: language pairs, credentials, travel, rush rates, and two-sided billing don’t fit general-purpose tools.
- FIA built IMP, the Interpretation Management Platform. It now runs our scheduling, dispatch, billing, interpreter pay, and VRI.
- It’s text-first because a Deaf owner built it. Nothing in it assumes you’ll “just call.”
- Other agencies can use it now. Nothing changes for FIA clients.
The three stages every agency goes through
Talk to ten agency owners and you’ll hear the same story in a different order. This is the version we lived.
One tab per month, interpreter availability from memory, invoices built by hand. Breaks at the first double-booking, or the first month you can’t reconcile what you billed against what you paid.
An appointment tool built for salons, clinics, or field service, bent to fit interpreting. It will book a slot. It has no idea what a language pair is.
Powerful, and built for national language companies. Priced that way too. In more than one case the vendor runs its own interpreter marketplace alongside the software you’re paying for.
When I started FIA in 2016, scheduling was a spreadsheet. Requests came in by email and phone, I matched them to interpreters from memory, and billing happened at the end of the month with a calculator and a lot of patience. It worked because we were small. It stopped working the moment we weren’t.
The generic scheduler that came next would book an appointment. It would not know that a Deaf patient needs two interpreters for a three-hour surgical consult, that one of our interpreters holds a legal credential and one doesn’t, or that the same job pays the interpreter one rate and bills the hospital another. So we kept a second system for all of that, and re-keyed everything between the two.
The enterprise platforms solve those problems. They also assume you have a procurement department, and more than one of them is run by a company that also sells interpreting, which means your supplier is competing with you for the same jobs.
The hidden cost at every stage is re-keying.
Every time a job moves between two systems by hand, someone retypes the date, the address, the language, the rate, and the interpreter. That is where double-bookings, wrong rates, and missed confirmations come from. The tool that removes re-keying is the tool that removes most agency errors.
What interpreting agency software actually has to do
This is the list we wish someone had handed us at stage one. It isn’t a feature wish list; every row is a place we lost money or credibility before we had it.
| Requirement | Why it matters | Why generic tools miss it |
|---|---|---|
| Language pair on every job | ASL, Spanish, and Dari are not the same resource | Schedulers have “services,” not languages |
| Credentials that gate assignment | Court and medical settings need specific qualifications | No concept of refusing an offer to an unqualified resource |
| Broadcasting, not just assigning | Agencies offer one job to several interpreters and take the first accept | One resource per slot is the whole model |
| Conflict detection with travel buffer | Finishing in Frederick at 2:00 means you can’t start in Baltimore at 2:15 | Overlap is checked by clock time only |
| Two-sided rates | What you bill and what you pay are different numbers with different rules | One price per service |
| Rate cards per client | Hospital, school district, and law firm each have a contract | The coordinator has to remember them |
| Invoicing that talks to accounting | Batch to QuickBooks, online payment, void and reissue cleanly | Export a CSV and hope |
| Portals for interpreters and clients | Interpreters accept from a phone; clients see invoices without emailing you | Staff-side only |
| Remote interpreting built in | VRI and OPI are part of every agency’s offer now | Separate vendor, separate login, separate bill |
| Text-first operation | Deaf owners, Deaf interpreters, Deaf clients, and coordinators at 11 p.m. | “Just call” is assumed everywhere |
Why we built our own
That last row is the one that tipped us. As a Deaf business owner, I don’t run my agency by phone. Every tool we tried assumed the coordinator would call the interpreter, call the client, or call support. I needed something that worked entirely in text, for me and for the Deaf interpreters and clients we serve, and nobody was going to build that for us.
So in 2025 I started building it inside FIA, on our own jobs. No outside requirements document, just the work in front of us every day. Today that system is called IMP, the Interpretation Management Platform, and it runs everything at FIA. It came together in the order the pain showed up:
Scheduling and broadcasting. The job, the language pair, the credential requirement, and a broadcast to qualified interpreters ranked by language, certification, and location, re-broadcasting until someone accepts.
Billing. Rate cards per client with rush, after-hours, and cancellation rules. Interpreter pay tracked separately. Charges calculate when the job is marked complete.
Portals. Interpreters accept, decline, set availability, see earnings, and get directions from a phone. Clients request, view history, and pay invoices online.
Remote calling. Video remote and phone interpreting moved inside the platform instead of through a third-party service.
AI where the bottleneck was. When the after-hours phone line was the problem, we built an AI receptionist that answers, takes the request in plain conversation, and drops a drafted job into the queue for a human to confirm. Urgent calls escalate to a person by text. When a hospital needed overnight dispatch, we built that.
None of it came from a feature roadmap. It came from a Tuesday when something went wrong.
What changed at FIA
| Before | After |
|---|---|
| Requests by phone and email, re-typed into the schedule | Portal, email, and phone requests land in one queue |
| Text interpreters one at a time until someone says yes | One broadcast to every qualified interpreter; first accept books it |
| Look up the client’s contract, calculate rush by hand | Rate card applies itself; billing finalizes when the job closes |
| Build invoices in a separate accounting tool | Batch invoices to QuickBooks; clients pay online |
| Interpreter pay in another spreadsheet | Tracked per job, separate from client billing, with pay schedules |
| Double-bookings discovered the morning of | Blocked at scheduling time, with travel buffer |
| VRI and phone interpreting through an outside vendor | Built into the same platform clients schedule from |
What manual work actually costs
A coordinator who re-keys 20 jobs a day at three minutes each spends an hour a day typing things a second time. Over a year that’s roughly 250 hours, before counting the errors re-keying produces. That hour is the difference between a coordinator who is scheduling and one who is firefighting.
What this means for clients and interpreters
Less than you might think, and that’s the point. The service is the same. The tooling underneath is different.
Request the way you always have, or through your portal. See job history and invoices, pay online, and connect to VRI from the same account you schedule from.
Accept or decline from your phone in one tap. Set availability, see what you’ve earned, and get directions to the on-site job. Fewer confirmation gaps, because nothing is re-typed.
Stopped re-typing. Fewer hand-offs means fewer mistakes, and a double-booking is blocked before it happens instead of discovered the morning of.
A 7 a.m. hospital job, before and after. The request arrives by voicemail at 9 p.m.
- Before: the coordinator hears the message at 7 a.m., calls three interpreters, reaches one at 7:40. The interpreter arrives late.
- After: the AI receptionist takes the request at 9 p.m. and drafts the job. The coordinator confirms it from her phone. The broadcast goes to every qualified interpreter within range, and the first accept is booked by 9:20 p.m.
- The hospital has a confirmation before the night shift changes.
Why we’re sharing it
Over the past year, other agency owners kept asking what we use. Small and mid-sized interpreting agencies, ASL and spoken language, healthcare and legal and education, all have the same operational shape and the same frustrations we had. There was no reason to keep the answer to ourselves.
So IMP is now available to other agencies. Flat monthly pricing, unlimited interpreters, and no marketplace competing with you for your own jobs, because we’re an agency and we wouldn’t accept that either. FIA is still customer number one and will stay that way. Every change we make, we make because we needed it first.
Ten questions to ask any agency software vendor
Whether you look at IMP or anything else, these would have saved us two of the three stages.
Does every job carry a language pair and required credentials, and will the system refuse to offer a job to an interpreter who doesn’t qualify?
Can I broadcast one job to several interpreters and book the first accept?
Does it detect conflicts with travel time, not just overlapping clock times?
Are client billing and interpreter pay separate numbers with separate rules?
Does it push invoices to my accounting system and let clients pay online?
Do interpreters and clients get their own portals, usable from a phone?
Is VRI or OPI built in, or am I paying a second vendor?
Can my whole team, including Deaf staff, run it without a phone call?
Does the vendor also run an interpreter marketplace that competes with me?
Is the price flat, or does it grow every time I add an interpreter?
Frequently asked questions
Does this change anything about FIA’s interpreting services?
No. Same interpreters, same coverage, same standards. The platform is the tooling behind the service, not the service.
Is IMP only for ASL agencies?
No. FIA provides ASL and 60+ spoken languages, and IMP was built for that mix from day one. It works for spoken-language, ASL, and blended agencies.
Does FIA see other agencies’ data on IMP?
No. Every agency’s data is isolated. FIA is a tenant on the platform like any other agency would be.
Do clients have to do anything differently?
No. Request the way you always have. If you’d like portal access to see invoices and history online, ask us and we’ll set it up.
I run an agency. How do I learn more?
Read how IMP came to be or see plans and pricing. You can also reach me through FIA. I’m happy to compare notes, in writing, anytime.
Need interpreters? Nothing about reaching us has changed.
Frederick Interpreting Agency is Deaf-owned and based in Maryland. ASL and 60+ spoken languages, on-site, virtual, VRI, and phone, across Maryland, DC, and Virginia, and remote nationwide. Same team, same standards, better tools underneath.
Ethan Kramer is the founder of Frederick Interpreting Agency and the creator of IMP. He works entirely in text and is happy to compare notes with other agency owners in writing. Reach him through FIA’s contact page.