Ringlop voice and chat agent for medical professionals
Budget: $1,500 – $3,000 USD
## Freelance Mission Brief – Build AI Agent Platform Backend + Web App (Ringlop)
**Type:** Freelance / Fixed Scope
**Budget:** \$2000
**Timeline:** 10–14 days
**Start:** ASAP
**Reporting To:** William, Founder of Ringlop
**Location:** Remote
**Ownership:** 100% of the codebase, infrastructure, and prompts belong to Ringlop
**Required Experience:** You must have previously built AI voice agents, web apps with secure user portals, backend automation flows, CRM/API syncs, and multi-channel integration systems
---
### Overview
**Ringlop** is an AI-powered growth assistant that replaces human setters and closers for MedSpas. Our product doesn’t just talk — it **books**, **qualifies**, **follows up**, **answers customer queries**, and **updates CRM records** across multiple channels, including:
* Phone (via voice agent)
* WhatsApp
* Instagram DMs
* Email
* Google Calendar and Gmail
Clients will access a clean web interface where they can monitor their agents, connect or disconnect channels, and track performance across time, channels, and agent instances.
You are building the **backend and admin-facing tooling** that allows Ringlop to launch custom, high-performing agents per client — one per business entity — while the frontend webapp provides client visibility and control.
---
## Scope of Work
### 1. Web Application (Frontend + Auth)
* Build a full **authenticated client dashboard**
* Each client sees:
* Their business name, active agent(s), and linked integrations
* Agent status (online, paused, inactive)
* Channel connections (WhatsApp, IG DMs, Gmail, Calendar, phone)
* Performance dashboard (calls answered, meetings booked, leads followed up, per channel and per agent)
* Ability to pause/reactivate agent per channel
* Admin users (us) can:
* Manually activate the agent
* Edit agent configuration
* View and override all client data
---
### 2. Backend Logic (Controlled Onboarding & Agent Creation)
* The onboarding form is not visible to the public or inside the client app
* It is used **by Ringlop (you)** post-demo **to collect structured data to build the custom agent**
* The backend must:
* Store all form data (Supabase preferred)
* Parse uploaded files (scripts, menus, FAQs)
* Crawl the client’s website for relevant info
* Build a **custom prompt** from the data
* Trigger agent generation via Vapi (voice) and chat stubs (WhatsApp, IG)
* Save all prompt versions and training inputs per agent
* After the form is submitted, the agent is inactive until validated and launched manually by Ringlop (admin user)
---
### 3. Voice Agent Routing (Inbound & Outbound)
**Inbound Calls:**
* Provision Twilio virtual number per client
* Clients forward their business number to Ringlop’s number
* AI answers using Vapi with client-specific prompt
**Outbound Calls:**
* Triggered by new leads, DMs, form fills, etc.
* Backend automatically initiates outbound call from Vapi
* Agent uses the prompt and logs the result
* Outcome (e.g., booked, no answer, callback scheduled) is tracked
---
### 4. Integration Hub (Connect/Disconnect Channels)
Clients must be able to securely connect and manage:
* WhatsApp Business
* Instagram DM (via Meta API)
* Gmail
* Google Calendar
* Calendly (optional)
* Phone (via forwarding to Ringlop number)
Integration flow:
* Ringlop creates and activates the agent
* Client logs into dashboard, sees their available channels
* Client connects relevant channels securely (OAuth or tokens)
* Once connected, Ringlop AI starts working live on those channels
---
### 5. CRM Integration + API Hooks
* Create internal API endpoints to push/pull from client CRMs
* At minimum, support webhook logic to:
* Create / update contacts
* Add call or message logs
* Change lead stage (e.g., contacted, booked, no-show)
* Provide a clear internal API doc so Ringlop can build connectors to tools like HubSpot, Pipedrive, GoHighLevel, or any webhook-compatible CRM
---
### 6. Security & Compliance (Must-Have)
* Encrypt all personal and integration data (OAuth tokens, phone numbers, names)
* Store call logs, transcripts, and message logs securely
* Agent first-message must include legal opt-in or disclaimer (text provided)
* Follow HIPAA-style and GDPR-style security practices:
* Scoped access roles
* Encrypted storage
* No direct exposure of PII in logs or LLM outputs
---
## Required Experience
You must have already delivered:
* Voice AI products using Vapi, Twilio, or similar
* Secure multi-step onboarding or admin tooling for AI apps
* Web apps with user authentication, dashboards, and API-based integration control
* Backend automation and prompt-generation logic with OpenAI or Claude
* CRM sync and webhook-based data syncing
---
## Suggested Stack
* Frontend: Next.js + Tailwind
* Backend: Node.js or Python (FastAPI okay)
* Database: Supabase
* Voice: Vapi
* AI: OpenAI or Claude
* Telephony: Twilio
* File Parsing: Langchain
* Web Crawling: Playwright / unstructured.io
* Integration Auth: OAuth2 / token handling
---
## Success Criteria
* The Ringlop team can onboard a new client via internal form → generate a fully trained agent
* After internal validation, client logs into their web dashboard
* Client sees their agent(s), connects desired channels, and the AI begins working live
* Clients can monitor activity and performance across time, channels, and agents
* All data is secure, and all systems are scalable
---
## To Apply
Please send:
* Links to voice/chat/automation tools you’ve built
* Confirmation you accept the \$2000 fixed fee
* Your availability and timeline
* Any suggestions you have to simplify or speed up delivery
No agencies. No beginners. Just experienced builders who know how to ship.
**Type:** Freelance / Fixed Scope
**Budget:** \$2000
**Timeline:** 10–14 days
**Start:** ASAP
**Reporting To:** William, Founder of Ringlop
**Location:** Remote
**Ownership:** 100% of the codebase, infrastructure, and prompts belong to Ringlop
**Required Experience:** You must have previously built AI voice agents, web apps with secure user portals, backend automation flows, CRM/API syncs, and multi-channel integration systems
---
### Overview
**Ringlop** is an AI-powered growth assistant that replaces human setters and closers for MedSpas. Our product doesn’t just talk — it **books**, **qualifies**, **follows up**, **answers customer queries**, and **updates CRM records** across multiple channels, including:
* Phone (via voice agent)
* Instagram DMs
* Google Calendar and Gmail
Clients will access a clean web interface where they can monitor their agents, connect or disconnect channels, and track performance across time, channels, and agent instances.
You are building the **backend and admin-facing tooling** that allows Ringlop to launch custom, high-performing agents per client — one per business entity — while the frontend webapp provides client visibility and control.
---
## Scope of Work
### 1. Web Application (Frontend + Auth)
* Build a full **authenticated client dashboard**
* Each client sees:
* Their business name, active agent(s), and linked integrations
* Agent status (online, paused, inactive)
* Channel connections (WhatsApp, IG DMs, Gmail, Calendar, phone)
* Performance dashboard (calls answered, meetings booked, leads followed up, per channel and per agent)
* Ability to pause/reactivate agent per channel
* Admin users (us) can:
* Manually activate the agent
* Edit agent configuration
* View and override all client data
---
### 2. Backend Logic (Controlled Onboarding & Agent Creation)
* The onboarding form is not visible to the public or inside the client app
* It is used **by Ringlop (you)** post-demo **to collect structured data to build the custom agent**
* The backend must:
* Store all form data (Supabase preferred)
* Parse uploaded files (scripts, menus, FAQs)
* Crawl the client’s website for relevant info
* Build a **custom prompt** from the data
* Trigger agent generation via Vapi (voice) and chat stubs (WhatsApp, IG)
* Save all prompt versions and training inputs per agent
* After the form is submitted, the agent is inactive until validated and launched manually by Ringlop (admin user)
---
### 3. Voice Agent Routing (Inbound & Outbound)
**Inbound Calls:**
* Provision Twilio virtual number per client
* Clients forward their business number to Ringlop’s number
* AI answers using Vapi with client-specific prompt
**Outbound Calls:**
* Triggered by new leads, DMs, form fills, etc.
* Backend automatically initiates outbound call from Vapi
* Agent uses the prompt and logs the result
* Outcome (e.g., booked, no answer, callback scheduled) is tracked
---
### 4. Integration Hub (Connect/Disconnect Channels)
Clients must be able to securely connect and manage:
* WhatsApp Business
* Instagram DM (via Meta API)
* Gmail
* Google Calendar
* Calendly (optional)
* Phone (via forwarding to Ringlop number)
Integration flow:
* Ringlop creates and activates the agent
* Client logs into dashboard, sees their available channels
* Client connects relevant channels securely (OAuth or tokens)
* Once connected, Ringlop AI starts working live on those channels
---
### 5. CRM Integration + API Hooks
* Create internal API endpoints to push/pull from client CRMs
* At minimum, support webhook logic to:
* Create / update contacts
* Add call or message logs
* Change lead stage (e.g., contacted, booked, no-show)
* Provide a clear internal API doc so Ringlop can build connectors to tools like HubSpot, Pipedrive, GoHighLevel, or any webhook-compatible CRM
---
### 6. Security & Compliance (Must-Have)
* Encrypt all personal and integration data (OAuth tokens, phone numbers, names)
* Store call logs, transcripts, and message logs securely
* Agent first-message must include legal opt-in or disclaimer (text provided)
* Follow HIPAA-style and GDPR-style security practices:
* Scoped access roles
* Encrypted storage
* No direct exposure of PII in logs or LLM outputs
---
## Required Experience
You must have already delivered:
* Voice AI products using Vapi, Twilio, or similar
* Secure multi-step onboarding or admin tooling for AI apps
* Web apps with user authentication, dashboards, and API-based integration control
* Backend automation and prompt-generation logic with OpenAI or Claude
* CRM sync and webhook-based data syncing
---
## Suggested Stack
* Frontend: Next.js + Tailwind
* Backend: Node.js or Python (FastAPI okay)
* Database: Supabase
* Voice: Vapi
* AI: OpenAI or Claude
* Telephony: Twilio
* File Parsing: Langchain
* Web Crawling: Playwright / unstructured.io
* Integration Auth: OAuth2 / token handling
---
## Success Criteria
* The Ringlop team can onboard a new client via internal form → generate a fully trained agent
* After internal validation, client logs into their web dashboard
* Client sees their agent(s), connects desired channels, and the AI begins working live
* Clients can monitor activity and performance across time, channels, and agents
* All data is secure, and all systems are scalable
---
## To Apply
Please send:
* Links to voice/chat/automation tools you’ve built
* Confirmation you accept the \$2000 fixed fee
* Your availability and timeline
* Any suggestions you have to simplify or speed up delivery
No agencies. No beginners. Just experienced builders who know how to ship.