One foundation, as many worlds as clients
Sphæra gives each of your clients a complete, rigorously partitioned online service — search, connected data, automation, identity, integrations, cockpit — without rebuilding the infrastructure every time.
The problem
Most platforms patch these concerns together by hand — orchestration, catch-up scripts, monitoring, human memory. Sphæra makes them reliable by design, from the foundation up, and reusable for every client.
Available today
Every capability is live. They combine freely to fit each client's needs.
Instant full-text search, autocomplete, relevance ranking, filters and live-count facets. An editable model describes the catalog: the same engine indexes any content structured by that model.
The same engine indexes any content described by a model — describe the catalog and it adapts without being rebuilt — from a thousand records to more than 4 million items, searched in tens of milliseconds.
AvailableBring data from your external tools — accounting, CRM, catalogs, files, APIs — into the client's world: real-time events via signed webhooks and Zapier, or imports from your sources, ready to be searched, displayed and automated.
Available today: signed webhooks, Zapier and imports. Continuously managed connectors remain an industrialization step.
AvailableThe engine that runs the work: submit a task and it schedules it — now, later or recurring — runs it (including code in a hardened, ephemeral container), then logs every run. Submission and tracking interfaces included.
No loss, a single instance in charge at a time, with a complete log.
AvailableThe secure front door: key-based login, signed access token, validation on protected calls, refresh and logout — down to rejecting invalid keys.
"Who is this caller, are they allowed in?" — solved.
AvailableTwo-way connectivity with Zapier: receive events from thousands of apps (with signature verification) and trigger actions back in those same tools. Credentials stay platform secrets.
The SaaS ecosystem — thousands of apps — without custom glue.
AvailableA dashboard dedicated to each client: the catalog of their services, their tasks and their runs. Deliberately read-only — a window, never a control panel. One application, configured per client at launch.
A client's dashboard, partitioned from all the others.
AvailableA fast memory space services share to cache results instead of recomputing everything. Deliberately "best-effort": if it's unavailable, the service keeps working, simply without the speed-up.
Extra speed, with no point of fragility.
AvailableA real-time window on the running system: which services are running, how they depend on one another, streamed live to the browser and drawn as a living graph.
See what's running live, without guessing.
AvailableThe client brings their own code through a simple contract — "one input, one result" (pricing engine, quote calculation, business rule…). The task engine automatically routes the relevant jobs to it, in that client's isolated world.
A quote, a price, a business rule — your logic, in the client's world.
AvailableIsolation, without the cost
One client = one isolated world that carries its own logic and into which the platform services it's entitled to are projected.
Data, search index, tasks, cache and identities are partitioned — separated from every other client, while sharing the same underlying infrastructure. The benefit fits on one line:
Per-client isolation, without paying for per-client infrastructure.
What already runs
How a client is assembled
It gets its business module and its isolated world. The platform's capabilities are shared products, not shared instances: each client receives a copy deployed in its own world.
Access works by projection, not by permissions: a service is available because it was deployed in the client's world. Requesting a service that wasn't granted produces a clear refusal — never a half-working state.
Two clients want to collaborate, exchange or share information? An explicit tunnel can be opened between their worlds — passing only what they choose, and nothing else. The connection is a deliberate choice, never a default.
What makes Sphæra different
These guarantees are exercised independently of client products: failures, restarts, work redistribution, temporary absence and a component's end of life are verified as platform laws, not as business special cases.
A restart is normal, an absence is a valid state, a failure is a signal you observe — not an incident to hide.
Work distributes itself, deterministically, with no central "brain" whose failure stops everything. Fewer breaking points.
For a given job, a single instance is responsible for it — no duplicates, no costly negotiation between servers.
Adding a client or a vertical search means differentiating an existing model, not starting from scratch. The cost of a new client drops as the platform grows.
Roadmap
These initiatives are scoped — in progress or planned. They turn an already-operational platform into a fully industrialized offering.
| Initiative | What it brings | Status |
|---|---|---|
| Billing & quotas | Per-client usage metering and aggregation for billing. The cost signals already exist; per-client accumulation unlocks pay-as-you-go and tiers. | Planned |
| Edge authorization control | Explicit refusals ("not authorized") and quotas enforced at public entry points. Makes access control explicit. | Planned |
| Multi-region | Regional variants (global, Canada, EU) already started on the search side — to be generalized for data residency and latency. | Started |
| Self-service portal | Today opening a client is operator-assisted; tomorrow the client provisions itself. | Planned |
| Packaged vertical search | Search "stamps" itself by domain; turning it into a catalog of ready-to-use verticals, fed by connected sources and kept up to date. | Started |
| Inter-client tunnels | Open an explicit connection between two worlds to share chosen information — never by default, always by choice. | Planned |
| Enriched cockpit | Live per-service health view, in the client dashboard. | Planned |
Offer structure
Built from what's actually deliverable. Cells marked roadmap arrive gradually, often reserved for higher tiers.
| Starter | Pro | Enterprise | |
|---|---|---|---|
| Isolated world per client | ✓ | ✓ | ✓ |
| Intelligent search | ✓ | ✓ | ✓ |
| Task automation | Basic | Recurring + isolated code | Unlimited + custom logic |
| Identity & access | ✓ | ✓ | ✓ + edge authorization / quotas (roadmap) |
| Zapier integrations | — | ✓ | ✓ |
| Client cockpit | ✓ | ✓ | ✓ + live health (roadmap) |
| Custom business logic | — | — | ✓ |
| Multi-region / residency | — | — | roadmap |
| Support & SLA | Community | Standard | Dedicated |
In short
Today, Sphæra delivers isolated SaaS worlds, one per client — each with its own logic, plus search, connected data, automation, memory, integrations and cockpit.
The next industrialization steps — billing, edge authorization, self-service, inter-client tunnels, multi-region — are already scoped. Want a world for each of your clients?