How is Melaya Assistant different from a provider chatbot?

A provider chatbot primarily answers inside its own interface. Melaya Assistant works across the connectors, context, pipelines, and paired devices you authorize, while preserving model choice, human approval, and an auditable execution trail.

Can an enterprise use its own models and infrastructure?

Yes. Teams can use supported cloud providers, enterprise-hosted endpoints, CLI subscription runtimes, or local models through Ollama and LM Studio. Model choice is not tied to the workspace or the interface.

How does Melaya protect enterprise data?

Connector credentials, conversation history, and device sessions are tenant isolated. The assistant sees only explicitly selected connectors and context, sensitive actions can require human approval, and execution is recorded for review.

Can Melaya Assistant operate apps on a phone?

Yes, on a paired Android device. It can read and operate approved apps through Device Control even when those apps have no automation API, subject to the device allowlist and approval policy.

What are Safe, Payments only, and Autonomy modes?

Safe mode asks for approval on consequential writes. Payments only lets routine requested actions proceed while retaining payment approvals. Autonomy executes the user's requested automation within configured permissions and platform policy.

Can I see what the assistant used and what it cost?

Yes. Melaya records provider rounds, tool calls, outcomes, token usage, attributable cost, approvals, and execution spans so operators can inspect what happened instead of relying on a chat transcript alone.

02 / 03

Isang assistant.Lahat ng system. Kontrolado.

Makipag-usap nang natural sa business apps, documents, agent pipelines, at Android phone mo. Piliin ang model, itakda ang permissions, aprubahan ang mahalaga, at siyasatin ang bawat action.

← Bumalik sa tahanan

Melaya AssistantGoverned at handa

Safe modeModel mo3 scoped systems

Aling overdue invoices ang kailangang asikasuhin, at ano ang susunod?

May nakita akong tatlong overdue invoices sa Odoo. Itinugma ko ang accounts sa recent email history at inihanda ang susunod na actions para sa review.

Odoo · invoicesreadZoho Mail · historyreadFollow-up draftsapproval required

Magtanong sa buong workspace…

ScopedKung ano lang ang pinili mo

MeasuredTokens at cost bawat round

TraceableTools, outcomes, approvals

PortableCloud, local, o server mo

Alam ng chatbot ang sasabihin. Alam ng Melaya kung ano ang puwedeng gawin.

Kapaki-pakinabang ang provider chatbots. Kailangan ng enterprise ng operating layer: malinaw na access, verified actions, model choice, cost controls, at ebidensiya pagkatapos ng usapan.

Enterprise requirement

Provider chatbot

Melaya Assistant

Business context

Provider chatbot

Files at text na kinopya sa chat

Melaya Assistant

Static context at selected live systems

Action

Provider chatbot

Advice o provider tools

Melaya Assistant

Scoped connectors, pipelines, at phone tools

Control

Provider chatbot

Prompt instructions

Melaya Assistant

Enforced modes at editable approvals

Model strategy

Provider chatbot

Isang provider catalog

Melaya Assistant

Cloud, enterprise, CLI, at local

Accountability

Provider chatbot

Chat transcript

Melaya Assistant

Provider rounds, tool traces, outcomes, at cost

Apat na tanong na dapat sagutin ng enterprise assistant.

Saan napupunta ang data? Magkano ang gastos? Ano mismo ang nangyari? Puwede bang baguhin ang deployment? Ginagawang visible at configurable ng Melaya ang bawat sagot.

Privacy by architecture

Tenant-isolated ang history, credentials, connector access, at device commands. Tanging context at systems na pinili para sa conversation ang natatanggap ng assistant.

Scoped access, encrypted secrets, isolated execution

Governable na cost

I-route ang bawat workload sa tamang model. Gumamit ng cloud APIs, CLI subscriptions, o local models. Provider-aware limits ang pumipigil sa biglang malaking bill.

Round-level tokens, cost, bounded context

Observability, hindi tiwala lang

Makita sa iisang trace ang provider calls, tool arguments, results, approvals, at final outcomes. Nasusuri ang bawat claimed action laban sa ebidensiya.

Replayable traces at truthful outcomes

Deployment na walang lock-in

Panatilihin ang interface at governance habang nagpapalit ng model o runtime. Melaya Cloud, enterprise server, supported CLI, o local.

Model choice bawat conversation at workload

Tanungin ang business, hindi isa-isang app.

Piliin ang systems na kailangan at magsalita nang natural. Kayang iugnay ng assistant ang impormasyon, magmungkahi ng next step, at ipa-review ang sensitive writes.

Explicit connector scope

Hindi awtomatikong exposed ang connected service. Pinipili ng user ang personal o project connectors bawat conversation.

Server-side ang credentials

Tool schemas at results lang ang nakikita ng model, hindi ang raw secret.

Mula tanong hanggang workflow

Ang matagumpay na request ay puwedeng gawing reusable pipeline na may parehong tools, permissions, at approvals.

Melaya AssistantConversation scope

Selected systems lamang ang available

Ibigay ang brief. Panatilihin ang thread. Iangkop sa model.

Iniaangkla ng static context ang policies at goals. Pinapanatili ng model-aware memory ang useful history nang hindi ino-overflow ang window gamit ang screenshots, tool results, o mahabang chat.

STATIC CONTEXT

Mga fact na hindi dapat magbago

Customer success policy.pdf

Escalation thresholds, approved language, account tiers, response targets…

PDF · TXT · MD

Conversation instructions
Magdagdag ng stable brief para sa tone, policy, role, at boundaries.
Document drops
Mag-attach ng PDF, TXT, o Markdown references.

MODEL-AWARE MEMORY

Mahabang usapan na walang blind overflow

Managed context budget

Totoong context windows
Umaangkop ang budget sa selected provider at model.
Buong tool rounds
Pinapanatili ng compaction ang call-result pairs para valid ang payload.
Redacted summaries
Puwedeng i-distill ang lumang context at i-redact ang sensitive patterns bago gamitin o i-store.

Piliin kung paano puwedeng mag-execute ang request.

User ang humihiling ng trabaho. Ang mode ang nagtatakda kung saan hihinto ang Melaya sa connectors, pipelines, at paired phone.

MAXIMUM REVIEW

Safe

I-pause ang consequential writes para sa malinaw at editable na human decision.

Review bago ang sensitive action

ROUTINE FLOW

Payments only

Ituloy ang requested routine actions habang gated pa rin ang payments at financial commitments.

Palaging kailangang aprubahan ang payment

REQUESTED AUTOMATION

Autonomy

I-execute ang task sa loob ng configured permissions, tools, app allowlists, at policy.

Patuloy na umiiral ang permissions at policy

Hindi pinalalawak ng modes ang access. Approval behavior lamang sa loob ng granted permissions ang binabago nito.

Bumuo ng totoong request bago mag-register.

Piliin ang connectors, mode, at task. Walang data na ina-access o ipinapadala ang preview. Pindutin ang Enter para magpatuloy.

Bagong governed conversationReady

Flexible model

Ano ang ipapagawa sa Melaya?

Magtanong sa selected systems o gumamit ng proven request.

Kausapin ang AI mo. Hayaan itong magtrabaho sa Android phone.

Para sa founders, creators, operators, at mobile teams, puwedeng ipagpatuloy ng Melaya Assistant ang requested task sa paired Android, kahit walang automation API ang app.

Totoong mobile interfaces

Magbukas ng allowed apps, basahin ang screen, mag-tap, mag-type, mag-swipe, at i-verify ang next state.

Voice kahit saan

Gamitin ang voice at mobile overlay para humiling, sumubaybay, at tumugon sa approvals.

On-device ang control

Android executor ang nagpapatupad ng app allowlist, publishing gates, stop controls, at watchdog.

Tuklasin ang Melaya Device Control

Melaya AssistantOnline ang phone

Ihanda ang replies sa mahahalagang comments ngayon at huminto bago mag-publish.

Nagtatrabaho sa phone mo

Binabasa ang allowed app at inihahanda ang susunod

Kausapin ang Melaya

Nananatili sa device ang publishing approval

Simpleng interface sa ibabaw ng governed runtime.

Natural ang usapan. Explicit, bounded, at inspectable ang execution.

01 · ASK

Sabihin ang outcome

Mag-type, magsalita, o mag-attach ng context.

02 · SCOPE

Piliin ang access

Piliin ang connectors, model, device, at mode.

03 · EXECUTE

Panoorin ang verified work

Sundan ang tools, results, phone steps, at approvals.

04 · PROVE

Suriin ang ebidensiya

Tingnan ang outcomes, cost, rounds, at buong trace.

Bago humawak ng totoong trabaho ang assistant.

Malinaw na sagot para sa security, platform, operations, at users.

Paano naiiba ang Melaya sa ChatGPT, Claude, o ibang provider chat?

Melaya ang governed operating layer sa paligid ng model. Selected systems lang ang ikinokonekta, gumagawa ito sa tools at Device Control, pinapanatili ang model choice, approvals, at execution evidence.

Puwede ba ang sariling model o infrastructure?

Oo. Gumamit ng cloud, enterprise endpoints, CLI subscriptions, Ollama, o LM Studio nang hindi binubuo muli ang experience at governance.

Exposed ba sa lahat ng conversation ang connected app?

Hindi. Pinipili ng users ang personal o project connectors bawat conversation at server-side ang credentials.

Ano ang puwedeng i-audit pagkatapos?

Provider rounds, calls, arguments, results, approvals, tokens, attributable cost, at final outcomes ay naka-record bilang ebidensiya.

Makokontrol ba ang phone apps na walang API?

Oo, sa paired Android. Device Control ang gumagawa sa allowed interfaces na may on-device allowlist at controls.

Ibig bang sabihin ng Autonomy ay unlimited access?

Hindi. User ang humihiling ng automation. Patuloy ang permissions, connector scope, app allowlists, safety controls, at policy.

Gawing governed execution ang conversation.

Magsimula nang libre gamit ang models at tools mo ngayon. Magdagdag ng connectors, mobile execution, at enterprise deployment kapag real na ang trabaho.

Magsimula sa Melaya Assistant Basahin ang documentation

Libreng simula · Built-in human control · Walang model lock-in

Sumali sa komunidad