A great deal of what looks like real-time AI is precomputed. As inference costs rose, doing the expensive work in advance and serving the result instantly went from optimisation to strategy. Precomputing is the term for it, and it covers a wide family: caches, materialised views, precomputed embeddings, prerendered pages, lookup tables replacing live calculation. Engineers building in this space need language for what they do, and this is the word itself on the .

Community energy cooperatives need names that sound solid rather than clever, and they’re growing across Europe and North America. This one sounds like it has been generating electricity since 1974. Both words are common and strong, and together they suggest essential infrastructure rather than a venture. That serves a generation, distribution, storage or industrial power business, and it serves a battery or UPS brand where the promise is literally keeping the power on.

Explaining throughput to a village committee deciding whether to fund a fibre scheme is a real communication problem. Megabits mean nothing; the word fast means everything. Community broadband projects face that conversation constantly, they exist across rural regions in a dozen countries, and they are organised as cooperatives and nonprofits on this extension. Bitspeed says the thing without the unit. Two plain words, an obvious join, no hyphen and no unusual spelling, reading as a real product rather than a coinage.

Batch workloads are where cloud spend gets wasted most spectacularly, and tooling that fixes it sells easily. That’s the commercial edge on an otherwise unglamorous category. Batch computing is permanent rather than fashionable. Jobs that run without anyone waiting: rendering, simulation, genomics pipelines, risk calculations, ETL, model training. It predates interactive computing and never went away. Every cloud provider sells a batch service, HPC centres run on the model, and the scheduling software behind it is a serious technical field with its own literature.

Media workloads are the heaviest thing most companies ever run, which makes this a demanding category and a well-paid one. Instance is the word cloud computing settled on for a running copy of something: EC2 instances, database instances, container instances. Media instances therefore reads as provisioned media infrastructure, whether that’s transcoding, streaming, rendering or storage delivered as spun-up capacity. The name speaks the right language to the buyer, who is an engineer or a technical operations lead rather than a marketer.

Cloud development environments, preview deployments and remote workspaces became standard practice at larger engineering organisations, and the tooling market around them is active. Code instances reads as exactly that: the ephemeral machine spun up for a branch, a review, a test run or a workspace. Instance is the cloud’s word for a running, provisioned thing, and the audience reads it without pause. The name suits a cloud IDE, a preview environment service, a CI infrastructure product, or a sandbox platform for teaching and demos.

Picking one of these is a decision teams get roughly one chance at. Data models are hard to migrate, operational characteristics only reveal themselves under load, and the vendor documentation is written to win evaluations rather than to inform them. An independent comparison resource would be consulted by every engineering team making that call, which is a steady and valuable audience. The underlying category is not going anywhere. Key-value storage sits beneath caching, session handling, configuration, feature flags and real-time state, and it has done for decades across successive generations of implementation.

The singular suits a product rather than a platform. One service, one clear function, which is how infrastructure components are actually sold. Instance is cloud vocabulary for a running, provisioned thing. Applied to media it reads as on-demand media compute: transcoding, streaming, rendering, packaging. That’s a demanding technical category, since media workloads are heavier than almost anything else companies run, and the services handling them sell to engineers who care about throughput and cost per hour rather than branding.

Choosing a database is a high-stakes decision made with poor information. A comparison resource in this space would be genuinely useful and this name positions one correctly. The landscape justifies it. SQL is fifty years old and going nowhere, while the ecosystem around it multiplied: NewSQL, distributed SQL, query engines over object storage, SQL layers atop non-relational systems, and a whole family of alternatives to particular databases. The alt- prefix does precise work by signalling adjacency rather than replacement, which is the honest position.

Engineers discount names that sound invented and trust ones that sound like something they’d find in a codebase. This is firmly the latter. A render cache stores the output of expensive rendering work so it doesn’t have to be redone, and the pattern appears everywhere: page generation, 3D and video production, map tiles, data visualisation. That makes it a strong name for an edge or CDN product, a rendering service, a static site infrastructure company, a 3D pipeline tool, or a performance consultancy.

Comparison content in this category performs unusually well, because the decision is high-stakes and the vendor material is useless for making it. The key-value database is foundational to data storage and has become commercially central as systems grew and latency budgets tightened. Caching layers, session stores, configuration systems and real-time state all run on the model. This is the exact term on the .com with no hyphen. For a technical audience, exact-match naming beats cleverness, because it demonstrates that the people behind the site share their vocabulary.

Concrete nouns make better technical brands than abstract ones, because engineers can picture what you’re describing. Both halves of this name are things you can see. A chassis is the frame everything else bolts onto; flow is the movement through it. Together they name the underlying structure carrying a process, which is a good description of what a workflow platform or a data pipeline actually is. The name suits a workflow engine, a data pipeline product, an orchestration platform, or an automation framework.

Almost nobody builds on a single model any more. Routing between providers, ensembling outputs, falling back when one fails, comparing results and managing cost across vendors has become structural to how AI systems work. The companies solving this are a real and fast-growing category of model gateways, routers and orchestration layers, and what the category lacks is a plain-English name. This is it. The name suits a routing or gateway product, a comparison and benchmarking service, an API aggregation platform, or a research resource tracking model performance.

Buyers evaluating this layer have a genuine information problem, since the products differ in ways that only become apparent in production. A comparison resource would be used constantly. The plural names the category rather than a single product, which is the right shape for a directory, a comparison site, or a publication covering the space. Model aggregation became structural to how AI systems are built: routing between providers, ensembling outputs, managing fallbacks, and comparing cost and quality across vendors.

The vendor-neutral form matters. This name covers every model family rather than tying a brand to one provider’s naming, which makes it more durable as the market shifts between providers. Inference is the operational half of machine learning and the half costing money every day. Training is a capital expense; inference is the running cost, and optimising it absorbs a great deal of current engineering effort. The name suits an inference service, an optimisation product, a hardware company, a benchmarking resource, or a cost management tool.

Organisations routinely discover their inference spend after it has already become a problem. Cost management is genuinely underserved here. Training gets the attention, but the recurring cost of running models in production determines whether an AI business works, which has made inference optimisation one of the most valuable specialisms in the field. The plural suits a service or volume business: many inferences, billed or measured, which is literally how the market works, priced per token and per call.

Engineers evaluating options search in precise vocabulary rather than in marketing language, which is why exact-term naming is an advantage for a developer-facing product. The singular suits a product rather than a service volume. One inference: the unit of work being optimised, measured and billed. Inference became the operational centre of the AI business. Running cost determines margin, and an entire supply chain formed around making it cheaper and faster: specialised hardware, quantisation, caching, batching, routing and serving frameworks.

The .net reads as native to database infrastructure rather than as a second choice, and it costs nothing with a technical audience. SQL outlasted every technology supposed to replace it, and the landscape around it multiplied rather than shrank: distributed SQL, query engines over object storage, SQL layers atop non-relational systems, and a steady supply of alternatives to specific incumbent databases. The alt- prefix signals adjacency rather than replacement, which is the accurate position.

Regulated industries, government, and any company with data it can’t send to a third party all want models running in their own environment. That compliance requirement is a market with real budgets and durable purchasing. Instance is the cloud’s word for a provisioned, running thing, so applied to language models this names a dedicated deployment rather than a shared API. The name suits a private deployment service, a managed model hosting product, an on-premise AI vendor, or a consultancy specialising in self-hosted inference.

Nobody plans to run six deployments. It happens: one for the EU, one for the regulated subsidiary, one for staging, one a team stood up and forgot, and suddenly there is a fleet nobody is managing. Fleet management for model deployments is an unsolved operational problem and a growing one, covering provisioning, version drift, cost attribution, access control and shutting down the instances everyone forgot about. The plural is the whole point of this address.