From Service Providers to Technology Innovators: A Reflection on LSPs in the AI Era

By Tyler Herron, CEO & Co-Founder, Eclipse Scheduling

I just finished reading Giovanna Patruno's excellent piece in MultiLingual Magazine, "From Service Providers to Technology Innovators: Repositioning LSPs in the AI era", and I haven't been able to stop thinking about it. The article puts language to something many of us in the language services industry have been feeling for years: the line between language services and language technology has effectively dissolved. The LSPs that thrive over the next decade will be the ones that can hold both identities at once — without losing the human expertise that made them valuable in the first place.

I want to share my perspective from inside an LSP-built software company, and explain where Eclipse fits into this shift.

The Inflection Point Is Real — and It Started Years Ago

Patruno argues that LSPs are at a "critical inflection point." From where I sit, that point has been quietly building since the rise of neural machine translation (MT) and accelerated dramatically with the public arrival of large language models. Enterprise buyers no longer ask whether you "use" technology. They ask which APIs you expose, how your platform integrates with their CMS, how your workflows route content by complexity, and how you measure outcomes.

In other words: clients are no longer buying translations. They are buying language infrastructure.

That reframing is hard for many legacy LSPs to accept, because it implies a different operating model, a different sales motion, and — most uncomfortably — a different self-image. But the data backs it up. According to Slator's market research, technology-enabled LSPs are consistently outperforming pure-service peers on both growth and valuation multiples. The CSA Research outlook on language AI points in the same direction.

Where I Agree With Patruno

A few of her points land especially hard for me:

1. "Technology adoption is now table stakes." Generic MT engines and off-the-shelf scheduling tools are commodities. Differentiation has to come from somewhere deeper.
2. Domain-specific AI beats generic AI. A healthcare LSP that fine-tunes models on medical terminology, payer documentation, and Section 1557 language access requirements will out-perform a horizontal MT provider every single time. The same is true in legal, education, and public-sector work.
3. Integration architecture is the new moat. The most defensible LSPs are the ones whose APIs and connectors quietly run inside their clients' EHRs, LMSs, CRMs, and contact centers.
4. Human-AI symbiosis — not replacement — is the winning model. The future is not "humans or machines." It is humans deciding and machines assisting.

That last point is the one I want to plant a flag on.

Where I Want to Push Further: Humans at the Front of Critical Decisions

Patruno frames "human-AI symbiosis" as a workflow design choice. I think it is bigger than that. For the kind of work LSPs handle — courtrooms, hospitals, IEP meetings, asylum interviews, on-set production, financial disclosures — AI should never be the final voice in a critical decision. The cost of a hallucinated medical instruction or a mistranslated legal term is not a customer-experience problem. It is a safety, civil-rights, and liability problem.

So at Eclipse, we have adopted a simple internal principle:

AI accelerates the work. Humans own the outcome.

That principle shapes every product decision we make. We will automate the parts of LSP operations that are repetitive, mechanical, and low-risk — matching, broadcasting, credential expiration tracking, invoicing, reporting, reminders, scheduling logic, document parsing. We will not automate the moments where a qualified human linguist, project manager, or coordinator needs to make a judgment call.

How Eclipse Is Helping LSPs Stay Competitive in the AI Boom

Eclipse was built by career interpreters and LSP operators who lived the chaos Patruno describes — disconnected systems, manual processes, and a constant pressure to "do more with less" while clients quietly raised the bar on what "modern" looks like. Our job as a language technology platform (LTP) is to give LSPs the same operational leverage that big enterprise software has given other industries, without forcing them to abandon what makes them differentiated.

Here is where we are leaning in:

1. Purpose-Built Infrastructure for LSPs

Generic scheduling and CRM tools were never designed for language services. They do not understand language pairs, certification expirations, modality (OPI, VRI, on-site), credentialing requirements, or the nuanced billing structures LSPs use. Eclipse's 16+ integrated modules — scheduling, billing, pay tracking, credentialing, reporting, messaging, on-demand, and projects — are designed around how LSPs actually operate.

2. AI Integrated Where It Genuinely Helps

We integrate AI for the right jobs:

• Intelligent provider matching that considers certification, specialty, location, performance history, and availability.
• Document parsing and intake automation for requests, POs, and credentialing docs.
• Smart reminders and confirmations that reduce no-shows.
• Reporting and analytics that surface fill-rate, utilization, and revenue insights without requiring a data team.

What we deliberately do not do: route a patient to a machine when a qualified human interpreter is needed, replace a human project manager's judgment on a complex localization project, or generate translations that go to end-users without a human linguist in the loop.

3. Open Integrations, Not a Walled Garden

Patruno is right that integration architecture is becoming the linguistic operating system. Eclipse exposes APIs and webhooks so LSPs can connect their operations to client EHRs, learning platforms, contact centers, and their own MT and TMS environments. We want to be the operations layer that makes the rest of an LSP's stack work harder — not another silo.

4. Compliance as a First-Class Feature

HIPAA, SOC 2, FERPA, and Section 1557 language access requirements aren't optional in healthcare, education, or public-sector work. We build for them by default so our customers can confidently pursue enterprise and government RFPs that would otherwise be out of reach.

5. Pricing Designed for Scale, Not Lock-In

LSPs grow in non-linear ways — a single contract can double monthly volume overnight. Eclipse's tiered, all-inclusive pricing gives LSPs unlimited users, all modules, and predictable scaling instead of per-seat penalties that punish growth.

A Note to Fellow LSP Owners

If you run an LSP and you read Patruno's article and felt a little nervous, I get it. We felt it too. The honest read is this: you do not have to become a technology company to win, but you do have to deploy technology with the conviction of one.

That means:

• Auditing every workflow and asking, "Is a human required for this decision, or are we just doing it manually out of habit?"
• Investing in tools that compound — platforms that get smarter as your data grows, not just point solutions that solve a single pain.
• Protecting the moments where your linguists' judgment, cultural fluency, and ethical reasoning are the entire product.
• Communicating to clients that your differentiator is not the AI — it is how thoughtfully you deploy it around human expertise.

The LSPs that internalize this will not be commoditized by AI. They will be the ones AI vendors and enterprise buyers come to when they need the language layer done right.

Closing Thought

Giovanna Patruno is correct that the most successful LSPs in the AI era will operate like technology companies. I'd add one phrase: technology companies that never forget who is actually doing the interpreting, the translating, and the cultural bridging on the other end of the API call.

At Eclipse, that is the line we walk every day. AI where it helps. Humans where it matters. Infrastructure that lets LSPs do both at scale.

If you want to see how Eclipse is helping LSPs operate like modern technology companies without giving up the human expertise that defines our industry, schedule a demo or explore our platform.

---

Further Reading

• Giovanna Patruno — From Service Providers to Technology Innovators (MultiLingual, Sept 2025)
• Slator 2026 Language Industry Index
• CSA Research — Language Services and Technology Outlook
• HHS Section 1557 — Language Access
• Eclipse — Platform Features · Modules · Pricing

— Tyler Herron, CEO & Co-Founder, Eclipse Scheduling