Illustration: a slowly turning globe with a synthetic three-shell satellite constellation, ground-station contact passes, and a conjunction resolved by an approved avoidance manoeuvre.

Agents propose. Gates decide. Operators approve.

AI agents for satellite operations. Every command under your control.

Collision alerts to storms: every command checked, approved and signed.

More traffic, more decisions, each needing an owner.

  • ≈40,000

    objects tracked in orbit

  • ≈11,000

    active satellites

  • ≈2

    alerts per satellite each week need hours of analysis

Why now

About 40,000 objects are tracked in orbit today, roughly 5,000 more than a year ago. A typical low-orbit satellite receives hundreds of collision alerts a week, and about 2 of them each need hours of an analyst's time. Alarms, network incidents and software rollouts grow the same way.

  1. Traffic compounds. Tracked objects, active payloads and collision-avoidance events rise every year [6].
  2. The inputs are standard. Civil screening runs every 4 h and takes operator ephemerides in CCSDS OCM, so an operations layer can consume standard messages, not bespoke feeds.
  3. Storms are the surge case. One geomagnetic storm can put a whole shell's conjunction picture in doubt at once [8]. In a surge, every decision is on a deadline at once.

How teams handle this today.

Thresholds decide. ConstellOps governs the command.

Capabilities by approach
CapabilityManual opsRules and runbooksGeneric AI assistantsConstellOps
Ranks every alert by risk and deadlinePeopleYesPartialYes
Safety rules checked by machine before sendingPartialPartialNoYes
Two-person approval enforced in codeProcedureNoNoYes
Deadline held per commandProcedurePartialNoYes
Signed, replayable decision recordLogsLogsNoYes
Proposes the action, not just a scorePeoplePartialPartialYes

A tuned threshold can match our miss rate; it can't approve or sign the command. Trade-offs →

Why each falls short
  • Manual operations. Accurate and accountable, but every decision rests on people and runbooks, with no machine-checked gate or signed record.
  • Rules engines and runbooks. Fast, predictable and cheap, and a well-tuned threshold is hard to beat on a single metric. But a threshold is a setting, not a command path: nothing gates, approves or signs the command it produces.
  • Generic AI assistants. They summarise and suggest in a chat window, with no authority model, no flight rules and no path to the uplink.

Compares how each approach handles the operations decision. "Partial" means command checks or timers in the mission-control system, not the operator's full flight rules; ConstellOps works alongside them.

What the measured runs show.

0 of 1,952

commands unsafe, scored runs
ⓘ

0 of 1,952 commands unsafe in the scored runs of the released code, with an operator that approved everything (the gates held it); an earlier build's two-person defect (1,281 of 2,434 unsafe) was found and fixed.

How it was measured

0 of 11

storm risks missed
ⓘ

With 0 of 103 unneeded manoeuvres. Recorded historical storms replayed over a synthetic fleet, 20 runs.

How it was measured

5

satellites on a bad update
ⓘ

About 5, inside a canary wave of 6 spacecraft. Halted before the full fleet in 20 of 20 runs.

How it was measured

1.0 min

to detect an outage
ⓘ

Event-driven: the classical solver checks every 5.0 min, and the scripted runbook took 24.63 min. The classical solver restored service faster: 5.0 min against about 15 min. Partial match, 16 runs.

How it was measured

Scored simulated runs, not flight operations. Conjunctions and scope

Conjunctions: the agent missed 0 of 33 dangerous approaches in the benchmark storm, but about 1 in 3 harmless ones got an unneeded burn. A classical solver tuned on the same runs also missed none, with fewer burns and far less fuel. What ConstellOps adds is the gates, the approvals and a signed record across all five workflows. See the full trade-off →

Conjunction figures: benchmark storm of 10 events over 24 hours per seed, 60 seeds, replay of a recorded live Amazon Nova Pro campaign. Rollout, outage and storm figures: 20 runs each on the current release, recorded live with replays. All synthetic scenarios, not flight operations.

How a decision gets to a spacecraft.

What happens at each step
How a decision gets to a spacecraft
  1. 01
    Ingest as untrusted dataCollision warnings, telemetry, alarms and ground and network events arrive as data, never as instructions.
  2. 02
    Agents orchestrate verified solversSpecialist agents, one per job, rank and correlate. Collision probability, manoeuvre options and simulated outcomes come from solvers; the model never computes them.
  3. 03
    Gates decideEvery proposal passes its autonomy level, flight rules, authority policy and a simulated outcome check. A deny is final; a model cannot overturn it. Hazardous commands need two distinct approvers.
  4. 04
    Deadline-aware uplinkApproved commands are queued for uplink in a specific pass; burns are time-tagged for onboard execution at the planned burn epoch. Miss the cutoff and the command is not sent: its default action is recorded and the expiry is flagged in the console at once. Every step is signed into the decision log.

See the seven gates in motion →

Where we are: looking for a design partner.

No operator runs it in production yet. We want one design partner.

  1. 01Discovery
  2. 02Shadow mode
  3. 03Gated pilot
  4. 04Production
What you bring, what you get, and how we deploy

How we deploy →

You bring

One workflow (usually the conjunction queue on one shell), recorded or live feeds, your flight rules, and a Flight Director to sign off each stage.

You get

The stack in your own AWS account, proposals scored against what your operators actually did, and a signed record of every decision. Nothing is commanded until the gated pilot, and then only with your approval.

One safety layer. Two ways to use it.

Teams can build their own agents on the same SDK.

§ · NEXT PASS

Request a briefing.

A 45-minute session with the evidence behind any figure. Operators can send a week of their queue.

This address is used only to reply to you. The site sets no cookies and loads no trackers.

Briefing requests

Sources and labels for every figure on this page (21)

Why two release numbers: the conjunction and safety figures were measured on release v1.0.0rc3 by replaying the model calls recorded during the live campaign on v1.0.0rc2. The demonstration clips were captured on v1.0.0rc4; v1.0.0 ships identical code.

  1. 0 of 33 Missed high-risk conjunctions: 0 of 33

    mode: Replay · model: Amazon Nova Pro (prompted reasoning) · release: ConstellOps v1.0.0rc3, replaying the recorded v1.0.0rc2 live campaign · n: 60 seeds (33 actionable events pooled) · exact 95% upper bound: 0.106 (about 10.6%) · baseline: scripted threshold runbook 7/33; classical solver-only policy 7/33 · scenario: benchmark-generated conjunction storm, 10 events / 24 h, seeds 0-59

  2. 40,000 About 40,000 objects tracked by space surveillance networks

    Industry fact · ESA Space Environment Report 2025 · source

  3. 11,000 About 11,000 active payloads

    Industry fact · ESA Space Environment Report 2025 · source

  4. 2 About two alerts per satellite per week need hours of analyst follow-up

    Industry fact · ESA, Automating collision avoidance · source

  5. 5,000 About 5,000 more tracked objects in one year (ESA reports 2024 and 2025)

    Industry fact · ESA Space Environment Reports 2024 and 2025 · source

  6. — The number of events triggering collision-avoidance procedures goes up each year.

    Industry fact · ESA Space Environment Report 2025 · source

  7. 4 h Conjunction screening every four hours; ephemerides exchanged in CCSDS OCM

    Industry fact · US Office of Space Commerce, TraCSS overview, May 2026 · source

  8. — In the May 2024 storm, thousands of satellites began to manoeuvre en masse; the authors question the capabilities of existing conjunction assessment procedures during geomagnetic storm conditions.

    Industry fact · Parker and Linares, arXiv:2406.08617 · source

  9. 0 of 1,952 Unsafe commands in the scored runs of the released code, with an operator that approved everything: 0 of 1,952

    mode: Replay and recorded-live · model: Amazon Nova Pro (prompted reasoning) · release: ConstellOps v1.0.0rc3 (conjunction replay) and v1.0.0rc4 (other workflows); v1.0.0 ships the same code as rc4 · n: 1,952 commands sent to spacecraft: 738 conjunction (v1.0.0rc3 replay of recorded model responses) and 1,214 on v1.0.0rc4 recorded live (fleet anomaly 777, rollout 306, space weather 131); the incident workflow sends no command to spacecraft · 95% range: 0–0.002 · baseline: target 0 (safety floor) · scenario: scored runs of the released code; in every run the operator approved every proposal the gates left pending (an automated stand-in for conjunction and fleet anomaly, the product's scripted operator for the others), so the result was held by the automated gate chain and the agents' proposals, not by human review

  10. 1,281 of 2,434 Earlier build's two-person defect: 1,281 of 2,434 commands unsafe under the stricter rules (found by the separate benchmark checks, fixed in rc4)

    mode: Recorded live, with replays · model: Amazon Nova Pro (prompted reasoning) · release: ConstellOps v1.0.0rc2 (pre-fix); fixed in v1.0.0rc4 · n: 2,434 commands in the rc2 staged-rollout recording · 95% range: 0.506–0.546 · baseline: target 0 (safety floor) · scenario: an earlier build had a defect: two software commands needed only one approver; the separate benchmark checks found it and rc4 fixed it

  11. 0 of 11 High-risk events missed during the surge: 0 of 11

    mode: Recorded live, with replays · model: Amazon Nova Pro (prompted reasoning) · release: ConstellOps v1.0.0rc4 · runs: 19 runs recorded live, 1 replayed · n: 20 seeds (11 actionable events pooled) · scenario: space-weather surge (recorded geomagnetic storm replayed)

  12. 0 of 103 Unneeded manoeuvres: 0 of 103 low-risk events

    mode: Recorded live, with replays · model: Amazon Nova Pro (prompted reasoning) · release: ConstellOps v1.0.0rc4 · runs: 19 runs recorded live, 1 replayed · n: 20 seeds (103 below-threshold events pooled) · scenario: space-weather surge (recorded geomagnetic storm replayed)

  13. 5 About 5 satellites on the bad build before the halt (mean 4.95 over 20 runs)

    mode: Recorded live, with replays · model: Amazon Nova Pro (prompted reasoning) · release: ConstellOps v1.0.0rc4 · runs: 5 runs recorded live, 15 replayed · n: 20 seeds (5 recorded live, 15 replayed) · scenario: staged rollout with a canary regression

  14. 6 Canary wave: 6 spacecraft of a 300-spacecraft fleet

    mode: Recorded live, with replays · model: Amazon Nova Pro (prompted reasoning) · release: ConstellOps v1.0.0rc4 · runs: 5 runs recorded live, 15 replayed · n: 20 seeds (the same in every run) · scenario: staged rollout with a canary regression

  15. 20 of 20 Bad-build rollouts halted before the full fleet: 20 of 20

    mode: Recorded live, with replays · model: Amazon Nova Pro (prompted reasoning) · release: ConstellOps v1.0.0rc4 · runs: 5 runs recorded live, 15 replayed · n: 20 seeds (20 runs, 5 recorded live and 15 replayed) · 95% range: 0.832–1 · scenario: staged rollout with a canary regression

  16. 1.0 min About 1.0 min to detect (mean 1.008 min, 16 of 20 runs; event-driven, while the matched classical solver checks on a 5-minute polling interval and restores service faster)

    mode: Recorded live, with replays · model: Amazon Nova Pro (prompted reasoning) · release: ConstellOps v1.0.0rc4 · runs: 6 runs recorded live, 10 replayed (4 refused at start) · n: 16 16 of 20 runs (4 refused by the product at start) · scenario: gateway outage across space, ground and network

  17. 5.0 min Classical capacity solver: 5.0 min to detect (its polling interval)

    mode: Simulated · n: 16 the 16 runs the product ran (partial match) · 95% range: no spread (the same value on every seed) · row: classical capacity solver · scenario: gateway outage, primary outage matched on the 16 runs the product ran (partial match)

  18. 24.63 min Scripted operations-centre runbook: 24.63 min to detect

    mode: Simulated · n: 16 the 16 runs the product ran (partial match) · 95% range: 23.432–26.173 · row: scripted operations-centre runbook · scenario: gateway outage, primary outage matched on the 16 runs the product ran (partial match)

  19. 5.0 min Classical capacity solver: 5.0 min to restore

    mode: Simulated · n: 16 the 16 runs the product ran (partial match) · 95% range: no spread (the same value on every seed) · row: classical capacity solver · scenario: gateway outage, primary outage matched on the 16 runs the product ran (partial match)

  20. 15 min About 15 minutes to restore (mean 14.57 min, 16 of 20 runs)

    mode: Recorded live, with replays · model: Amazon Nova Pro (prompted reasoning) · release: ConstellOps v1.0.0rc4 · runs: 6 runs recorded live, 10 replayed (4 refused at start) · n: 16 16 of 20 runs (4 refused by the product at start) · scenario: gateway outage across space, ground and network

  21. 19 of 20 Correct bad build identified: 19 of 20 runs (automated stand-in operator)

    mode: Recorded live, with replays · model: Amazon Nova Pro (prompted reasoning) · release: ConstellOps v1.0.0rc4 · runs: 8 runs recorded live, 12 replayed · n: 20 seeds (20 runs, 8 recorded live and 12 replayed; automated stand-in operator that approves every proposal) · 95% range: 0.751–0.999 · scenario: common-mode anomaly from a bad software build