Cloud-native Architecture
Context
The situation before the platform
Why a recruitment organisation needed one intelligent ecosystem instead of a CRM, a spreadsheet and an inbox.
| The business |
A recruitment organisation in the United States, running candidate pipelines, client relationships and consultant workflows across a hiring operation that needed to scale. |
| The starting point |
Candidate data, communication history, recruitment workflows and hiring activity sat across emails, spreadsheets, CRMs and collaboration tools. Centralised tracking and operational visibility were extremely difficult. |
| The trigger |
Consultants were spending excessive time on repetitive administrative work — note-taking, drafting communication, coordinating scheduling and managing workflow — instead of candidate and client relationships. |
| What they wanted |
One unified digital ecosystem centralising candidate pipelines, consultant communication, scheduling, approvals, call tracking and hiring activity, with real-time visibility across the whole recruitment lifecycle. |
| Constraints |
Enterprise scalability had to hold as recruitment volume grew · AI output had to be reliable enough to trust in documentation and client-facing communication · workflow consistency had to survive consultants who each worked a different way. |
System
What it runs at today
The platform as delivered, live and scaling across web and mobile.
Recruitment visibility
40%
Improvement in recruitment visibility once pipelines, activity and communication resolved to one dashboard
AI capabilities
5
Call transcription, process automation, content generation, pipeline management and conversational search
Engagement length
7+
Months, as a product engineering partnership through build and scaling
Team size
8+
Design, frontend, backend, AI integration and QA across six delivery phases
The engineering problem
Six gaps that shaped the build from day one
Not vague pain points — the specific operational inefficiencies in a fragmented recruitment operation, each paired with what we did about it.
0
1
Recruitment workflows spread across disconnected platforms
Teams managed candidate communication, scheduling, approvals, notes and hiring activity across multiple systems that did not talk to each other. The result was workflow confusion and lost operational efficiency at every handover point.
What we did
Centralised candidate pipelines, scheduling, approvals, communication management and recruitment activity into a single AI-powered operational dashboard covering the full hiring picture.
0
2
Recruitment visibility nobody could get to
Candidate data, communication history, workflows and hiring activity lived in emails, spreadsheets, CRMs and collaboration tools. Tracking a hire centrally, or seeing how the pipeline was really performing, was extremely difficult.
What we did
A centralised dashboard bringing pipelines, workflow statuses, consultant activity, approval processes and communication logs into one interface, so progress is monitored in real time without tool switching.
0
3
Every consultant ran the process differently
Different consultants followed varying operational workflows and hiring processes. That produced inconsistent recruitment practice, weaker collaboration between teams and lower overall operational productivity.
What we did
Automated recruitment workflows covering candidate tracking, interview scheduling, follow-up management and approvals, so the same sequence runs the same way regardless of who owns the role.
0
4
Candidate progress tracked by hand
Following candidate progress, interview stages, follow-ups, approvals and communication history manually added operational complexity and slowed recruitment execution — the more active roles there were, the worse it got.
What we did
Intelligent candidate pipeline management with stage monitoring, bottleneck identification and clear hiring progress, giving teams the coordination they were previously reconstructing by hand.
0
5
Operational knowledge that lived in people, not systems
Critical recruitment knowledge, consultant workflows and organisational insight stayed undocumented or scattered across systems, creating dependency on individuals and disconnected operations when they were unavailable.
What we did
An operational intelligence layer capturing recruitment knowledge, organisational workflows and consultant activity into AI-accessible systems, reachable through conversational natural-language search.
0
6
Consultant time going to admin rather than people
Note-taking, drafting communication, scheduling coordination and workflow management consumed the hours that should have gone into candidate and client relationships. Productivity was capped by paperwork, not by demand.
What we did
AI transcription with automated call summaries, plus AI-generated job descriptions, adverts, emails and outreach, so documentation and drafting stop being manual steps in the consultant's day.
Architecture
How it fits together
Simplified — the shape rather than every service.
0
1
Clients
React.js · TypeScript · Role-based dashboards
Centralised recruitment dashboards, candidate management interfaces and operational reporting built in React.js and TypeScript, with a role-based interface that works across web and mobile.
0
2
Gateway
Nest.js · API layer · Node.js · Candidate management APIs
A type-safe Nest.js backend exposes candidate management APIs and carries every client request, keeping one contract behind the dashboard, the automation engine and the AI services.
0
3
Services
AI workflow automation · OpenAI APIs · Email automation services · Communication APIs · Real-time notification systems
Recruitment automation services, AI workflow engines, communication systems and operational intelligence modules each stay their own concern, so automation can be extended without destabilising the rest.
0
4
Data & infra
PostgreSQL · TypeORM · Azure
PostgreSQL with TypeORM holds candidate, workflow and communication data, deployed on Azure cloud infrastructure with cloud-based synchronisation built for enterprise recruitment volume.
The AI layer was designed in from phase two rather than added later: scalable AI workflow architecture, centralised recruitment data and operational intelligence layers were specified together, which is why automation reaches every stage of the lifecycle instead of one feature.
Solutions
Six systems doing the actual work
Not a features list — the specific things we built behind every number above.
AI-powered recruitment dashboard
Candidate pipelines, scheduling workflows, approvals, communication management and recruitment activity centralised into one operational dashboard for complete hiring visibility.
AI call transcripts & summaries
Interviews, client calls and internal discussions become structured transcripts with concise summaries, key highlights, action points and follow-up insights.
Automated recruitment workflows
Candidate tracking, interview scheduling, follow-up management, approvals and recruitment operations run automatically, cutting manual workload and accelerating hiring cycles.
AI-generated recruitment content
Job descriptions, recruitment adverts, candidate outreach, emails and follow-up communication generated from context, keeping messaging consistent as well as fast.
Operational intelligence platform
A centralised intelligence layer capturing recruitment knowledge, organisational workflows and consultant activity, and turning live activity into actionable insight.
Candidate pipeline management
Centralised candidate tracking with stage monitoring and bottleneck detection, improving recruitment visibility, workflow monitoring and operational decision-making.
Key features
What the platform does day to day
Five AI-driven capabilities, each closing one of the gaps identified above.
| Capability | Runs | Refresh | What it does |
|---|---|---|---|
| Automated call transcription & insights | Per call | On call end | Converts calls into accurate transcripts with instant AI-generated summaries |
| Smart recruitment process automation | Workflow-triggered | Continuous | Streamlines scheduling, follow-ups, approvals and workflows to accelerate hiring |
| AI-based content creation | On request | On demand | Generates job descriptions, emails and consultant communication instantly |
| Intelligent candidate pipeline management | Consultant-managed | Real time | Tracks candidate progress and provides centralised hiring visibility |
| Conversational global AI search | Natural language | On query | Answers questions across recruitment data with context-aware responses |
Integrations
How the moving parts plug in
Meetings, email, notifications and AI reach the platform through one API layer rather than sitting beside it as separate tools.
Connected systems
Zoom
Call workflows and recordings
Microsoft Teams
Collaboration and meeting workflows
OpenAI APIs
Transcription, summaries, content
Zoom Info
Contact and company data
↓
through one Nest.js API layer ↓
Platform integration layer
Nest.js API layer
One contract for every client
AI workflow automation
Scheduling, follow-ups, approvals
Notifications & email automation
Real-time alerts and outreach
↓
into the core services ↓
Core services
Candidate pipelines
Communication
Scheduling & approvals
Call tracking
Operational intelligence
Because call tracking, communication and pipeline data resolve against the same services, a consultant finishing an interview, filing its summary and moving the candidate to the next stage are three steps in one dashboard rather than three separate systems.
Security
What protects candidate and operational data
The platform holds candidate records, call transcripts and consultant communication, so access control and reliability were part of the architecture rather than a later pass.
Role-based access
A role-based interface means recruiters, managers and stakeholders each reach the pipelines, approvals and reporting their role requires and nothing beyond it.
Azure cloud infrastructure
Deployment on Azure with cloud-based synchronisation frameworks, optimised during the final phase for enterprise-level recruitment scalability.
Reliable data management
PostgreSQL with TypeORM gives type-safe, consistent handling of candidate records, communication history and workflow state across the platform.
Validated AI responses
AI response testing, automation testing and workflow validation ran before release, so generated summaries and content behave predictably in live recruitment use.
Process
How we got there
Five stages, starting with how recruitment work actually happens rather than a feature list.
0
1
Discover
Recruitment workflow analysis, stakeholder interviews, operational audits and communication process evaluations to identify inefficiencies and automation opportunities.
0
2
Define
Consultant workflows, candidate journeys, operational dependencies and AI automation requirements mapped, then prioritised into workflow automation, AI summaries and visibility features.
0
3
Design
Centralised recruitment dashboards, AI workflow interfaces, communication management systems and operational intelligence experiences designed around consultant productivity.
0
4
Develop
AI-powered workflow systems, candidate tracking modules, operational intelligence layers and automated communication services built on React.js, Nest.js and Azure.
0
5
Validate
Workflow validation, AI response testing, automation testing, usability analysis, performance optimisation and quality assurance ahead of cloud deployment.
Business impact
What changed for the business
Beyond the headline numbers, three things recruitment teams noticed first.
Consultants back on relationships
With note-taking, drafting and scheduling coordination automated, consultant time moved from administrative overhead to candidate and client engagement.
Hiring that moves faster
Automated tracking, scheduling, follow-ups and approvals took manual steps out of the pipeline, so roles progress instead of waiting on coordination.
One consistent way of working
Replacing disconnected systems with a centralised platform strengthened workflow consistency across consultants and made organisational knowledge accessible again.
Commercial outcome
What the engineering choices are worth in operating terms
Every headline number traces back to a specific decision, not a vague platform effect.
| Engineering decision | Operating outcome | Measured effect |
|---|---|---|
| AI transcription & automated summaries | Post-call documentation is generated instead of written up by hand | 60% less manual note-taking |
| Automated scheduling, follow-ups & approvals | Recruitment steps advance without consultants chasing coordination | 45% faster workflows |
| Centralised dashboard & pipeline management | Pipelines, activity and communication readable in one place, in real time | 40% better visibility |
| AI content generation & operational intelligence | Drafting and knowledge lookup stop consuming consultant hours | 51% more productivity |
Stack
What it's built on
The actual technologies, not feature names with icons attached.
Frontend
- React.js
- TypeScript
Backend
- Nest.js
- Node.js
- TypeORM
Data & infra
- PostgreSQL
- Azure
- AI workflow automation
Integrations
- OpenAI APIs
- Microsoft Teams
- Zoom
- Zoom Info
- Communication APIs
- Email automation services
- Real-time notification systems
Full case study
Get the complete write-up as a PDF
The same content on this page, plus the extended module breakdown and delivery phases, in a single document you can share internally.
Full AI workflow and service architecture
Six delivery phases, from discovery to deployment
Automation and integration decisions in detail
Download the case study
No spam • unsubscribe anytime • we’re here when you need us
More work
Other platform builds
Hospitality
Every stay, extraordinary.
Hospitality group transformed guest experiences through IoT-enabled room automation, reducing service requests by 58%
Hospitality Group transformed Guest Experiences through IoT-enabled Room Automation, reducing Service Requests by 58%
IoT - Hospitality
Read the case study →