The Resolv platform

The IT operations platform run by AI agents.

A complete ITSM and AI-operations platform: specialist agents, every channel your staff use, ticketing, assets, change management and analytics — engineered to enterprise DevOps standards, and simple enough for anyone to use.

21

platform capabilities

3

editions — SME to Government

2

languages, native RTL

1

audit trail under it all

The AI core

A team of specialists, not one bot

An orchestrator plans every request and routes it to the right specialist agent — each with its own tools, its own knowledge domain, and defined authority. Plain answers on the surface; structured, auditable reasoning underneath.

front line · triage

Help Desk

Picks up every request, resolves the everyday majority, routes the rest.

identity · entra id

Active Directory

Accounts, groups, resets and revocations — identity handled properly.

licences · mailboxes

Microsoft 365

Mailboxes, licences, sharing and tenant hygiene, end to end.

devices · servers

Windows & Linux Diagnostics

Reads the logs, checks disks and services, proposes the fix.

write-ups · root cause

Documentation & RCA

Documents every fix and drafts the root-cause analysis nobody had time to write.

network · security · more

A growing roster

New specialists join the team as the platform grows — same orchestration, same guardrails.

All features

21 capabilities. One platform.

Everything below ships in a single platform — no add-on maze. Statuses are honest: what runs today, what is in development, and what belongs to the Ops Platform Edition.

01

AI Help Desk

Available now

The front line. Every request picked up the moment it arrives, triaged, and — for the routine majority — resolved on the spot from your runbooks and environment.

Instant pickup

No queue, no business hours. First answer in seconds.

Triage built in

Category, priority, language and sentiment on every request.

Grounded answers

From your knowledge — when it doesn’t know, it says so.

Clean escalation

What it can’t fix moves up with the full story attached.

02

Specialist AI agents

Available now

An orchestrator routes work to specialists — Active Directory, Microsoft 365, device diagnostics, documentation & RCA — each with its own tools, knowledge domain and defined authority.

Defined authority

Each agent can only do what its role allows. Nothing more.

Deep domain context

Identity questions go to the identity expert, not a generalist.

Structured handoffs

Agents pass typed, auditable state — never loose text.

A growing roster

New specialists join under the same orchestration and guardrails.

03

Knowledge & RAG

Available now

Three layers of knowledge — runbooks, documents, past resolutions — with health scoring that tells you what the AI knows and gap detection that tells you what it’s missing.

Cited answers

Every answer links the runbook or fix it came from.

Health scoring

A live score for how well-covered your environment is.

Gap detection

Unanswerable questions become a to-write list, automatically.

Self-improving

Every resolved ticket makes the next answer better.

04

Omnichannel intake

Available now

Microsoft Teams, email, web chat and voice — plus WhatsApp — feeding one conversation layer. Arabic dialects and English, detected per message.

One conversation

Same request, same context, whichever channel it came from.

Teams-native

A bot inside the tenant your staff already live in.

Email becomes tickets

Inbound mail is parsed, triaged and answered — no forwarding rules.

Bilingual by design

Arabic and English replies, native RTL, per-message detection.

05

Ticketing & unified inbox

In development

Humans and AI work the same queue. Every conversation is a first-class entity with SLA timers, ownership, and reply-back to the end user from one workspace.

One queue

AI-resolved and human-owned work, visible side by side.

SLA timers

Response and resolution clocks with breach alerts.

Reply from the inbox

Agents answer the end user directly — on the original channel.

Full context handoff

Transcript, attempts and a recommended fix on every escalation.

06

Assets & service catalogue

In development

Every device and licence on record, and a structured menu of IT services staff can request — so the AI acts with context and staff stop writing freeform messages.

Asset register

Devices, licences, warranties and owners in one inventory.

Context for the AI

“My laptop is slow” resolves against the actual laptop.

Service catalogue

Structured requests with the right approvals attached.

Lifecycle tracked

From order to assignment to return — nothing walks away.

07

Change management

Available now

Log, review, approve, schedule and close IT changes — the formal process NESA, CBUAE and DESC expect from regulated organisations, without the ceremony.

Approval chains

The right sign-offs, enforced — not remembered.

Change calendar

Every scheduled change on one timeline, conflicts visible.

Status-page sync

Planned work communicated to staff automatically.

Audit-ready record

Who approved what, when, and what changed — permanently.

08

Workflow automation

Available now

Describe a process in plain language and get a running workflow — triggers, AI steps, conditions and human approvals on a visual canvas.

Plain-language builder

A sentence becomes a canvas you can refine.

Real triggers

Ticket created, priority raised, schedule, webhook, sentiment.

Humans in the loop

Approval steps anywhere a person should decide.

Run history

Every execution logged, inspectable and re-runnable.

09

Analytics & CX Score

Available now

One composite score that tells leadership how IT is performing — with SLA, sentiment, cost and topic analytics underneath, and proactive outreach before staff even raise tickets.

CX Score

A single number for IT health, computed continuously.

Topic clustering

What staff actually struggle with, grouped automatically.

SLA & cost analytics

Where time and money go, per team and per topic.

Monthly report

A board-ready PDF, generated — not assembled.

10

AI quality engine

Available now

The AI monitors its own quality: simulated scenario suites, hallucination detection and confidence calibration gate what reaches your staff.

Simulation before staff

Hundreds of scenarios run against every change.

Hallucination flags

Ungrounded answers caught and quarantined.

Confidence calibration

Low confidence means escalate — never guess.

Continuous evals

Golden datasets and red-team suites, run on every release.

11

Agent copilot

Available now

Your human engineers get their own AI: drafted replies, suggested fixes, conversation summaries and one-click escalation calls — assistance, with the person in charge.

Drafted replies

A grounded draft waiting when the engineer opens the ticket.

Suggested fixes

What worked the last three times this problem appeared.

Instant summaries

A long conversation compressed to what matters.

One-click call

Escalate to a live call with the staff member instantly.

12

CRM & customer context

Available now

Contacts, companies and a pipeline view — synced with your CRM so every conversation knows who it’s talking to and what they’re worth to the business.

Contact directory

Every staff member and stakeholder, with history.

Company records

Org-level context attached to every conversation.

Pipeline view

A kanban of relationships for account owners.

CRM sync

Zoho lookup during conversations; more connectors on the way.

13

Self-service portal & status page

Available now

A public knowledge portal where staff help themselves, and a status page that answers “is it down?” before anyone opens a ticket.

Knowledge portal

Branded, searchable self-service for common answers.

Public status page

Live service health with incident history.

Incident comms

Planned maintenance and outages, announced once.

Deflection by design

The cheapest ticket is the one never raised.

14

White-label & branding

Available now

Your logo, your colours, your bot’s name. Staff experience their company’s IT department — Resolv is the engine underneath.

Full branding

Logos, colours and tone across chat, portal and email.

Named personas

Your bot, your name — in Arabic and English.

Branded reports

Monthly reports carry your identity, not ours.

Per-company config

Every tenant styled and configured independently.

15

SLA & proactive operations

Available now

SLA policies with live timers and breach alerts — plus an AI that watches for trouble and reaches out before staff raise tickets.

SLA policies

Response and resolution targets per priority and team.

Breach alerts

Warnings before the clock runs out, not after.

Proactive detection

Emerging issues spotted from patterns, not complaints.

Morning briefing

What happened overnight and what needs attention, daily.

16

Billing & cost governance

Available now

AED billing with UAE VAT handled correctly — and something rarer: every AI call metered, costed and capped per company.

AED-native billing

5% VAT, TRN on every invoice, Stripe subscriptions.

Per-call cost logging

Every model call metered — you see what AI actually costs.

Spend caps

Hard monthly caps per company. No surprise bills.

Usage metering

Consumption visible per tenant, per feature, per month.

17

API & integrations

Available now

A versioned REST API, webhooks, and connectors into the systems you already run — Microsoft 365, Entra ID, Jira, Zoho, RMM tooling.

REST API v1

Tickets, conversations and knowledge, programmable.

Signed webhooks

HMAC-verified events into your own systems.

Enterprise connectors

M365, Entra ID, Jira, Zoho — read and act, both ways.

Encrypted credentials

Per-company AES-256-GCM; never exposed to the model.

18

Memory & knowledge graph

Ops Platform Edition

Long-term organisational memory: entities, relationships and history linked in a knowledge graph, so agents reason over your organisation — not just the current message.

Organisational memory

People, devices, systems and their relationships, linked.

Session context

Short-term memory scoped to each conversation.

Cross-signal reasoning

“Third laptop complaint from finance this week” is a signal.

Tenant-scoped always

Memory never crosses company boundaries.

19

Approval center & policy engine

Ops Platform Edition

Every pending AI action in one governed queue — with a policy engine that decides what agents may do unaided, what needs sign-off, and what is forbidden.

One approval queue

Every pending action, one place, one click to decide.

Policy rules

Authority defined per agent, per action, per environment.

Auto-approve the safe

Routine actions flow; sensitive ones wait for a person.

Decision audit

Who approved, who denied, and why — recorded forever.

20

Automation marketplace

Ops Platform Edition

A curated library of workflow templates and connector packs — install proven automations in one click instead of building from scratch.

Template library

Onboarding, offboarding, access reviews — pre-built.

One-click install

Deploy a template, then tune it to your environment.

Connector packs

Bundled integrations for common stacks.

Curated & reviewed

Every template reviewed before it reaches the library.

21

Ops command center & sovereign deployment

Ops Platform Edition

The enterprise/government cockpit: incident center, investigation workspace, per-agent workspaces and the full admin console — deployable inside your own perimeter.

Incident center

Detection to post-mortem in one governed flow.

Investigation workspace

Deep-dive tooling for the hard problems.

Sovereign deployment

Kubernetes + Terraform, in your data centre or sovereign cloud.

Zero-Trust program

IAM, isolation, threat model and compliance frameworks.

Built for two

Staff just ask. IT stays in control.

Plain language on the surface for every employee. Operator-grade depth underneath for the people responsible.

For your staff

Ask like you’d ask a colleague.

No portal training, no ticket forms, no jargon. Staff describe the problem in their own words, in Arabic or English, on the channel they already use.

  • Answers in seconds, any hour
  • Teams, email, web chat — no new app
  • Arabic or English, detected automatically
  • Requests from a simple service menu
  • Told honestly when a human takes over

For the IT team

Command-center depth.

A unified inbox, formal change control, a real API and an audit trail that survives scrutiny — the operator tooling an IT pro expects.

  • One queue for AI and human work, with SLAs
  • Change management with enforced approvals
  • Every AI action attributed in an append-only log
  • REST API, webhooks and enterprise connectors
  • RBAC, SSO and per-company policy control

Who it serves

One platform. Three editions.

The same AI core and the same suite — packaged for how much control, compliance and infrastructure you need.

SME

Your whole IT department, managed

For companies without a big IT team: Resolv is the help desk, the engineer and the paperwork.

  • Live in days — we onboard your knowledge for you
  • Arabic and English support on the channels staff already use
  • Onboarding and offboarding run end to end
  • Simple monthly pricing, no infrastructure to own
Enterprise

An AI workforce inside your ITSM

For IT organisations that need depth, control and compliance on top of speed.

  • Specialist agents for identity, M365, devices and RCA
  • Formal change management for regulated industries
  • SSO, role-based access and append-only audit
  • CX Score, SLA analytics and a developer API
  • Integrations: Microsoft 365, Entra ID, Jira, RMM
Government

Sovereign, auditable, in your perimeter

The Ops Platform Edition: the full system, deployed and governed on your terms.

  • On-premise or sovereign-cloud deployment — Kubernetes and Terraform
  • Zero-Trust architecture with isolation enforced at the database
  • Every AI action policy-gated, human-approved and audited
  • Data residency and compliance-framework alignment
  • Red-teamed AI evaluation gates every release

The engine room

Enterprise DevOps underneath. Plain language on top.

You don’t need to know what Kubernetes is to use Resolv. Your platform team will be glad it’s there.

Event-driven microservices

Every module scales and fails independently — one slow part never takes the rest down.

Kubernetes + Terraform

Infrastructure defined as code — rebuilt identically in any region, or in your own data centre.

SLOs & observability

We measure our reliability the way we measure yours — and publish the status.

HA & disaster recovery

Replicated, backed up, rehearsed. Recovery is a runbook, not a scramble.

AI evaluation & red-teaming

Golden datasets, hallucination detection and adversarial tests gate every release.

Append-only audit

The record of what happened cannot be edited — by us, by you, or by the AI.

This page shows the full platform scope. Availability varies by edition and rollout stage — ask us for the current status of any capability.

See the pipeline run on your real tickets

A 30-minute demo on your actual scenarios. No commitment.

Live in days · Nothing to install · Every action audited