01.0
Agent-native workflows
Assign AI agents to translate, review, and sync content while keeping human review first-class.
The localization platform to launch globally in days
Purpose-built for localization in the AI era.
Designed for human-in-the-loop.
Reviewing the source copy
Supported LLM providers
A localization operating system in AI era. Built for modern teams with translation, review, sync, and quality control in one workflow.
02.0
Bring your LLMs. Keep your TMS.
Stay flexible across model providers and existing TMS platforms without redesigning your operating model.
03.0
Designed for release confidence
Carry shared context, evals, and regression checks through every sync so quality does not quietly drift.
01 Sources
Localization begins in the tools teams already use
Pull requests, launch requests, and campaign assets become structured localization work with source context attached from the start.
Sources
GitHub, Slack, Claude
GitHub
Pull request
Changed files
messages/en/pricing.json
docs/launch-checklist.mdx
locales/fr-FR/hero.json
3 locale updatesdrift ready
Repo changes become intake.
Pull request diffs arrive with file context intact.
Slack
Request
#launch-ops
New request
Need hero copy and pricing updates translated for the Friday campaign launch.
fr-FRde-DELaunch copy
Requests arrive with urgency.
Timing and locale targets stay attached from the start.
Anthropic
claude
mcp add --transport http hyperlocalise
https://hyperlocalise.com/mcp
Trigger translation
Claude can start Hyperlocalise work with locale scope and source context attached.
01.1 GitHub changes
01.2 Slack requests
01.3 Claude briefs
01.4 Context attached
02 Translate Task
Move translation work across agents, reviewers and systems
Assign agents to translation, review, and sync tasks while keeping human reviewers in the loop through the TMS your team already uses.
Translate Task
Localise Pricing Messages
Working ...
Assign to...
02.1 AI translation + review
02.2 TMS Agnostic
02.3 Regression Evals
02.4 Context Auto-discovery
03 Providers
Stay flexible across providers and platforms
Hyperlocalise sits between LLM providers and TMS tools, giving teams one workflow even as models, vendors, and review systems change underneath it.
Providers
Azure OpenAI
AWS Bedrock
LM Studio
Ollama
LILT AI
03.1 LLM-agnostic
03.2 TMS-agnostic
03.3 Provider switching
03.4 Workflow continuity
04 Evaluations
Review PRs and translation regressions before merge
Run GitHub Actions for localization drift, compare outputs, and keep release checks visible in pull requests before broken sync reaches production.
apps/web/messages/fr.json
review translation updates
3 strings changed, 1 translation issue called out before merge
Review required
OldNewDiff
41
41
{
42
- "pricing.trialCta": "Start your free trial",
43
- "pricing.noCard": "No credit card required",
44
- "pricing.launchMarkets": "Publish to 12 markets in minutes",
45
- "pricing.billingCycle": "Billing available monthly"
44
+ "pricing.launchMarkets": "Publiez sur 12 marches en quelques minutes",
Comment on line R44
Hyperlocalise requested changes
Missing accent changes the French noun
French requires “marchés” with an acute accent. Without it, the line reads like the verb “marches” instead of “markets” in customer-facing pricing copy.
Mention Hyperlocalise to patch the diff in-thread.
Hyperlocalise is replying
42
+ "pricing.trialCta": "Commencez votre essai gratuit",
43
+ "pricing.noCard": "Aucune carte de crédit requise",
45
+ "pricing.billingCycle": "Facturation disponible chaque mois"
46
46
}
04.1 Drift checks
04.2 Eval gates
04.3 PR visibility
04.4 Release blocking
05 Monitor
Understand translation quality at scale
Track locale health, review coverage, and eval history in one place so localization PMs and engineering teams can see what is safe to ship.
Locale ops observability
Release window activeShip confidence across eval health, review debt, and failure modes.
Ship-safe locales
92%
Review SLA
4.1h
Critical blockers
1
Release Readiness
Locale readiness heatmap
Quality
Review
Drift
Sync
Status
fr-FR
96
94
91
95
Ship-safe
de-DE
93
91
89
92
Ship-safe
es-ES
90
78
88
91
Review due
ja-JP
88
80
82
89
Review due
pt-BR
81
64
76
83
Blocked
ko-KR
94
92
90
94
Ship-safe
Eval Trend
Quality score over recent runs
92%
current ship-safe coverage
Coverage Mix
Review coverage by locale
AI drafted
Human reviewed
Blocked
Failure Modes
Issue breakdown
Ranked by impact across current release checks
50 total findings
05.1 Locale health
05.2 Eval history
05.3 Review coverage
05.4 Release readiness
Recent releases
Changelog
Product updates that tighten release confidence, reduce localization drift, and make sync runs easier to trust.
LatestApr 15, 2026
v1.4.4
Added ICU parsing, improved markdown validation flow, and reused existing translated entries to avoid redundant work.
ReleaseApr 10, 2026
v1.4.3
Shipped summary JSON output, lazy prompt handling, and check fixes to make release reporting easier to trust.
ReleaseApr 10, 2026
v1.4.2
Introduced progress checks and tightened locale flag handling, validation routing, and merged-locale deduplication.
ReleaseApr 9, 2026
v1.4.1
Added check --fix, AST/ICU split, severity tuning, and --quiet mode for cleaner CI output.