Act 2: Moving Pictures

It’s been three years with GenAI in our lives.

You may think we understand what AI can or cannot do (at least those in the Global North think they do). But think again. There are entire regions of the world that are playing catch-up to the AI-first reality. And there are always newcomers to our industry who need guidance to set them on the right path.

Ultimately, progress will look different to every team and professional. It won’t be linear, and it may come at a cost. It may require us to redefine ourselves. In our quest onward, no one should be left behind, right?

In this section, read the articles that explore the human side of the (r)evolution unfolding around us.

István Lengyel

Founder and CEO at BeLazy

István Lengyel is the Founder and CEO of BeLazy, a company working towards introducing exception-based translation project management. In this role, he analyzed and designed full automation with over ten TMSes and a variety of business management systems. He was also the Co-Founder of memoQ and spent many years conceptualizing and designing the system‘s concepts and functionality.

Hype has tunnel vision. The AI craze zeroed in on one thing: where to throw automatic text processing at problems. Everything else was ignored.

Now we‘re sobering up from our LLM hangover and realizing we can‘t just sprinkle AI magic dust on everything. Two reasons:

1. Humans still matter. (Yes, that‘s the optimistic take!)
2. Economics matter. If we deploy armies of agents with mountains of background material, won‘t it cost the same as just hiring human translators?

Reality check: you‘re not managing one technology anymore. You‘re managing a zoo of them, or someone else is doing it for you. Welcome to the era of technology vendor management! You need the right mix of providers — humans and machines — to cover all bases without breaking the bank.

Here‘s the paradox: While every conference preaches hourly fees, LLMs and MT are doubling down on cost-per-word pricing. TMSs keep pushing “managed words“ — reinforcing per-unit pricing while experts sell the opposite. Measuring the two and normalizing them is key to keeping sanity.

Build your own tech stack or pay the premium. Managed services cost more than DIY — that‘s universal. So here‘s your roadmap: think in three independent tracks. List all steps of your different workflows. By workflow I don’t mean what any TMS, business management system (BMS) or other system calls a workflow, it is what you are doing, from the moment something appears in a system — and your email or a link is as much a system as a customer portal — to the moment you will never touch this content again. Map every workflow step against each track, and suddenly the gaps become obvious.

The three tracks

1. Project flow: The step-by-step journey of your project. TMSs excel at getting documents in/out of bilingual formats. BMSs and task tools like JIRA handle the human prep work better. Some steps are manual (DTP), some are TMS-native (translation/review), and others are automated (segment locking based on MTQE). Once you‘re in a TMS, XLIFF export-import is your friend.
2. Modality: Human work, LLM, machine learning, traditional MT, or rule-based engineering — what are you applying and where? Both MT software and LLMs handle multiple file formats beautifully in their UIs, but LLM APIs? They are restricted. LLM providers added all these flashy features to impress your boss and leave the headaches to you.
3. Databases: Four big dynamic ones plus smaller static ones like style guides. The dynamic quartet? The translation repository (translation memory, corpus, collection of bilingual files — any of these concepts works), term base or glossary, client database, vendor database. TMSs excel at managing the first two, but oversimplify the latter two, while BMSs, spreadsheets, and CRMs are the usual go-to for client and vendor management. The translation repository is where final, approved translations get stored for two reasons: reusing them and using them to orient any of the above modalities. Don‘t forget that for efficient “personalization” of modality queries, only relevant information needs to be extracted and passed over. While TMSs generally store these databases, for such purposes, something smaller and more tailored may even be enough.

The integration reality

TMSs used to be the integration hub. The TMS providers know this and they charge accordingly. However, their solution isn‘t necessarily the best or only solution for you.

Enter middleware, orchestrators, and custom functions. Combine these with any TMS and you‘ll get good results. Mix technologies freely and you‘ll get better results. You can even use one TMS’s file filters and user interface while tapping into another TMS’s AI capabilities and connector framework — the possibilities are endless.

Integration truth: I‘ve never seen impossible integration, but I‘ve seen plenty of cases where integration costs exceeded user benefits.

Vendor myth-busting: Don‘t assume that one vendor’s own integrated technology will be integrated better than systems combined from multiple (trusted) parties. The bigger the range of capabilities, the more gaps you can find. Focus means efficiency. We live in the microservices age — every “unified” system uses hundreds of components from strangers.

Flip the script

Technology changes daily. The tech-to-human cost ratio keeps growing. So, stop asking, “What’s possible?”

Start asking: “Which technologies and people optimize my speed-cost-quality triangle?”

That‘s where the real answers live.

Read the full 132-page Global Ambitions: (R)Evolution in Motion publication featuring vital perspectives from 31 industry leaders on the ongoing AI-spurred (r)evolution.

Get your digital copy today