Halo · by Le Website
CCaaS platform · In production · Validated under stress testing

Your entire contact center,
on a single platform.

Halo unifies telephony, the full Meta suite (WhatsApp, Instagram, and Messenger), and ticket management in a single agent screen. Smart queues, order taking, live supervision, quality assurance, and executive reporting — built for mission-critical operations where every conversation counts.

0Concurrent agents validated under load
0Calls at the peak of stress testing
0Automated tests passing
0Uptime target
One system, your entire operation
WebRTC TelephonyWhatsApp Business APIInstagram DMMessengerTickets and queuesOrder takingLive supervisionQuality (QA)Outbound campaignsExecutive reporting
The platform in action

This is what operating with Halo looks like.

A real walkthrough of the supervision panel: today's KPIs, every agent's status, queues by department, and ticket management — all updating in real time over WebSockets, with no page refresh.

Supervision panel · live
halo · supervision
Everything it does

Eleven modules. One agent screen.

Every module is designed around the ticket as its backbone: calls, conversations, and orders always stay connected to a case, with full history and traceability.

Voice / PBX in the browser

WebRTC softphone with nothing to install: IVR queues, hold music, overflow, transfers, hold, call recording, and agent states (available, break, on call, after-call work).

💬

Meta omnichannel messaging

Direct integration with the full Meta suite (WhatsApp Business API, Instagram DM, and Messenger): a unified inbox shared across agents, a first-level bot with menus and data capture, approved templates, and outbound campaigns — every channel in the same queue and the same ticket.

Tickets, departments, and queues

Organizational departments with business hours, assignment schemes (round-robin, least busy, manual), automatic transitions, custom customer fields, SLA monitoring with supervisor alerts, and a complete stage workflow.

🛒

Order taking

A form built into the agent screen with phone-based screen-pop, a catalog with prices frozen at confirmation, draft → confirmed → dispatched states, dispatch agencies with capacity control, and CSV export.

Live supervision

A real-time management dashboard: calls in queue, agents by status with activity history, open tickets, SLA, and instant case reassignment.

🎧

Call monitoring

Supervisors can listen to a live call without being heard, whisper to the agent only (live coaching), barge into the conversation with the customer, or mute the agent — all from the browser.

↩️

Callback queue

No abandoned call is ever lost: missed calls feed a callback panel for agents, grouped by number, with attempt counts and wait times.

Quality (QA) and CSAT

Interaction scoring with configurable criteria, per-agent summaries, post-interaction satisfaction surveys, and quality metrics by period.

📊

Executive reporting

Historical reports by channel, department, and agent: volumes, response times, dispositions, minute-by-minute activity, Excel/CSV export, and an automatic executive report emailed every morning.

Campaigns and bulk upload

Create outbound campaigns from Excel/CSV with background processing, progress tracking, and rate-controlled WhatsApp sends.

AI at ticket close Included

Automatic conversation summaries, suggested dispositions, and sentiment analysis when each ticket closes, plus a copilot that drafts replies. Every plan includes a monthly AI operations allowance, expandable in blocks.

NewArrives via voice or chat
AssignedSmart queue
In progressAgent + order
ResolvedSummary and disposition
ClosedFull traceability
How it's built

An architecture designed to never go down.

A modular monolith with a dedicated satellite process for voice: telephony stays isolated from API deployments, and every piece has a single responsibility. Simple to operate, hard to knock down.

Agent browserNext.js · SIP.js softphone (WebRTC)
Digital channelsMeta (WhatsApp · Instagram · Messenger) · webhooks
Telephony carrierSIP trunk
▼   WSS · REST · webhooks   ▼
api — FastAPIAuth · tickets · omnichannel messaging · orders · reports · real-time WebSockets
voice-gatewayARI client: call events → agent state · originate · monitoring
Asterisk 20 LTSIVR · queues · recording · WebRTC (DTLS-SRTP)
▼   domain events (pub/sub)   ▼
PostgreSQL 16Source of truth · history · reporting
RedisPresence · pub/sub · job queues
workersCampaigns · SLA watchdog · executive report by email · WhatsApp sends · backups

Each process runs in its own container. An API deployment never interrupts a call in progress: voice lives on its own lifecycle.

01Decisions that deliver stability

  • Voice queues live in Asterisk (20 years in production), not reimplemented: IVR → queue → agent in the dialplan.
  • A single per-agent state machine reconciles system presence with telephony presence — one button, two systems in sync.
  • Hot state in Redis for real time; persistence in PostgreSQL for history and reporting.
  • Nothing slow lives in the request path: campaigns, reports, and bulk sends run in workers with retries.
  • Event-driven architecture with a transactional outbox: every business change lands in an auditable history that feeds real time and reporting — nothing gets lost, even through outages.

02Your own instance, fully isolated

  • Every customer gets a dedicated instance: your data never shares space with another company's.
  • Data isolation reinforced at the code layer too: every query ships with scoping built in.
  • Granular roles and permissions: voice agent, text agent, supervisor, administrator, and management.
  • Action auditing and full traceability for every case.
Robustness and security

Built for operations where failure is not an option.

Telephony is mission-critical: if the system goes down, the operation stops. That's why every layer of Halo is hardened for real production and validated under stress testing: 60 simultaneous calls with 80 connected users and peaks of 500 calls, with zero platform errors.

🛡

Encryption at every layer

TLS across all traffic, WebRTC with DTLS-SRTP for voice, and call recordings encrypted at rest. Customer data belongs to the customer.

🔑

Secure sessions with rotation

JWT authentication with short-lived access tokens and refresh tokens rotated on every use — a stolen session expires on its own, and agents never get logged out mid-shift.

🚫

Active protection against attacks

Protective middleware that bans scanners, a firewall that exposes only what's needed, and SIP hardened against brute force — toughened by real production experience.

💾

Data protection

Log rotation across every service — a full disk never takes down the database — plus automatic daily backups with 30-day retention available as an add-on.

📡

Real time that heals itself

If an agent loses connection, their tickets return to the queue automatically and their status shows up in supervision instantly. Nothing is left orphaned.

🧪

Engineering quality

Over 535 automated tests passing on the business logic, strict typing end to end (Python type hints + TypeScript strict), and versioned schema migrations.

Service model

We don't sell you software. We run your platform.

Halo is delivered as a managed service: we deploy it, back it up, and evolve it — your team just focuses on serving customers.

🚀

Full implementation

Setup of departments, queues, schedules, templates, and users to match your current operation — with assisted migration and a zero-downtime transition. Included with your annual contract.

🔧

Direct support from the team that built it

No middlemen, no tickets to an external call center: you talk directly to the engineers who build the platform.

📈

Continuous improvement

Product improvements reach your instance as part of the service. Custom integrations (ERP, courier, internal systems) are quoted separately — included in 24-month Enterprise contracts.

You've seen the platform.
Now let's talk numbers.

Simple, published pricing — no surprises on your invoice.

Per-seat plan

Every role has its price.
Admins are free, always.

Pay only for the seats your operation needs, in the roles you need. Ideal for teams of 5 to 25 seats.

💬

Text agent

$65/mo

WhatsApp, Instagram, and Messenger in the unified inbox, with templates and tickets.

Voice agent

$80/mo

Browser softphone, IVR queues, recording, screen-pop, and order taking. Text channels included too.

🎧

Supervisor

$90/mo

Live dashboard, reassignment, listen/whisper/barge on calls, QA, and full reporting.

Always included
⚙️

Admin

$0

Up to 2 admin users to configure departments, queues, templates, catalog, and users. Free, always.

Minimum billing: $300 USD/month · Annual contract (12 months). Combine seats however you like; if your mix adds up to less than $300, the minimum is billed. Initial implementation and training are included with your contract, and you can adjust your seat mix month to month throughout the year. The per-seat plan applies up to 25 seats; for larger operations, the Enterprise license is the better fit. Prices do not include applicable taxes.
Full telephony in the browser (IVR, queues, recording)
All 3 Meta digital channels: WhatsApp, Instagram, and Messenger
AI included: summaries, dispositions, and sentiment at ticket close
Order taking with catalog and dispatch agencies
Mobile app + Web App, nothing to install
20 GB of data storage included
Your company's own subdomain
Rubric-based QA and CSAT surveys
Standard support with AI assistance
Calculate your price

Build your team and see
your monthly bill instantly

Adjust the numbers to match your operation. No surprises: what you see is what you pay.

$0
Build your package · optional add-ons
$0
Your monthly bill
$300
USD/mo · AI and telephony included
With VAT (13%): $339/moPrices are shown before VAT; the tax is itemized on your invoice.
The $300/mo minimum billing applies
With more than 25 seats, the Enterprise license is the better deal ↓
For large operations

Enterprise isn't more seats.
It's a different level of service.

From ~25 seats on, your operation demands guarantees the per-seat plan doesn't include. That's why Enterprise is contracted differently: a flat monthly license, sized to your operation.

Enterprise · no per-seat pricing

Everything in the per-seat plan, plus contract-backed guarantees

What a large operation gets in writing, not on a handshake:

  • Priority support, direct with engineering: critical incidents with response times defined by contract
  • Dedicated infrastructure: your own isolated instance, with capacity reserved exclusively for your operation
  • Assigned account manager: a single point of contact who knows your operation — no tickets to an external call center
  • Custom integrations (ERP, courier, internal systems), included in 24-month contracts
  • Advanced onboarding: guided implementation, zero-downtime migration, and on-site training
  • Grow without renegotiating: your license adjusts itself to your operation, with expanded AI across your entire operation
Flat monthly license
sized to your operation
Let's talk about your operation

One monthly number sized to your operation — and if you grow, it adjusts on its own, no renegotiation.

Optional add-ons

Clear extras, clear prices

Nothing gets activated unless you ask for it, and every extra has a published price.

Add-onWhat's includedPrice
Daily backupsAutomatic daily backup of all your data with 30-day retention$19/mo
Additional storageExtra 25 GB blocks for conversations, recordings, and files$10/mo per block
Additional AI blockRaises the monthly allowance of summaries, dispositions, and analysis$20/mo per block
Telephony minutesSIP trunk with the carrier of your choice, managed by usCarrier rates + management
Initial implementationSetup of departments, queues, IVR, templates, and team trainingIncluded with your annual contract
Custom integrationsConnection to your ERP, courier, or internal systemsQuoted · included in 24-month Enterprise
Frequently asked questions

What we always get asked

And anything missing, we answer over WhatsApp the same day.

Because every customer gets their own managed instance: infrastructure, monitoring, updates, and direct support from engineers. The minimum guarantees that even the smallest team gets the same level of service as a large operation. At 5 or 6 seats, most teams clear the minimum naturally.

An active user with an agent role (text or voice) or a supervisor role. Admin users (up to 2) don't count and aren't billed. You can change your role mix whenever you like; the change shows up on your next invoice.

Yes. Every plan includes a monthly allowance of AI operations: automatic summaries at ticket close, suggested dispositions, sentiment analysis, and QA reviewer assistance. If your operation uses more, you can add AI blocks for $20/mo. Competitors charge for this as a per-agent add-on.

You'll want to move to the Enterprise license: a flat monthly license sized to your operation, with expanded AI and integrations included in 24-month contracts. If your team grows, the license adjusts automatically — no renegotiations, no surprises. Migration is instant: it's the same platform, only your plan changes.

Because it isn't the same product as the per-seat plan. The per-seat plan is self-service with standard support; Enterprise includes response times guaranteed in writing, dedicated infrastructure, an account manager, and custom integrations — and that gets sized to each operation. That's why it's quoted as a flat monthly license: one number that covers your entire operation. Tell us about your case and we'll deliver your proposal the same day.

Yes: the contract is annual (12-month minimum), and in exchange, initial implementation and team training are included at no cost. During the contract you can adjust your seat mix month to month as your operation changes. At the end of the year it renews automatically, or you can cancel with 30 days' notice. Custom integrations are included at no cost in 24-month Enterprise contracts.

In your own instance, with full per-customer isolation, TLS across all traffic, and encrypted recordings. Your data is yours: if you ever leave, we hand all of it back in standard formats.

Next step

See it running on your own operation.

Let's set up a live Halo demo with your real workflows: your company's departments, queues, numbers, and use cases — and we'll deliver your exact proposal the same day.

Francisco González · fjgonzalez@le-website.com · +503 7931-1403