AI / Web Application
AI Resume Platform
A web platform that parses resumes into structured data, matches candidates to roles with an explainable score, and generates tailored documents.
- Next.js
- Python
- FastAPI
- OpenAI
- +1
View Case Study
SaaS / AI
Smart CRM
AI-Assisted Customer Relationship Platform
A multi-tenant CRM with AI assistance built into the record, summarising history and surfacing the next action for every account.
- Next.js
- TypeScript
- PostgreSQL
- OpenAI
- Node.js
Overview
Project Overview
Built for an operator whose process did not fit any off-the-shelf CRM, Smart CRM models their actual pipeline stages and permissions rather than approximating them. AI sits inside the record, not beside it: every account carries a rolling summary of its history, a suggested next action with the reasoning shown, and deduplication that runs continuously rather than as an occasional cleanup.
Services provided
- Custom Software Development
- Web Application Development
- CRM & API Integrations
- Client
- Confidential — multi-brand services operator
- Industry
- Professional Services
- Timeline
- 16 weeks
- Year
- 2024
- Platforms
- Web application
- Team
- 3 engineers, 1 designer, 1 delivery lead
Challenge
The Challenge
Their process spanned three tools and a large spreadsheet. Nobody trusted the data, and onboarding a new team member took weeks of tribal knowledge transfer.
Customer history was split across a generic CRM, a shared inbox and a spreadsheet of exceptions.
Duplicate records accumulated faster than anyone could clean them.
Reporting required manual export and reconciliation every month.
The existing CRM could not express their permission model, so everyone effectively saw everything.
Solution
What We Built
We modelled their real pipeline rather than a generic one, enforced isolation in the database, and put AI where it removes reading work instead of where it looks impressive.
Designed a schema around their actual stages, exceptions and role boundaries.
Enforced tenant and role isolation with row-level security, removing a whole class of access bugs.
Generated account summaries on write so reps stop reading months of history before a call.
Ran duplicate detection continuously at the point of creation with a human confirmation step.
Replaced the monthly export ritual with live reporting built on the same source of truth.
Capabilities
Key Features
What Smart CRM does in day-to-day use.
Multi-tenant by design
Tenant isolation enforced at the database level with row-level security rather than in application code.
Rolling account summaries
Each record carries an up-to-date narrative of its history, regenerated as new activity lands.
Next-action suggestions
Recommended follow-ups with the signals behind them shown, so a rep can disagree knowingly.
Continuous deduplication
Embedding-based matching flags likely duplicates as they are created, not months later.
Configurable pipelines
Stages, fields and permissions defined per tenant without forking the codebase.
Activity timeline
Calls, emails and system events unified into one chronological view per account.
Stack
Technology Stack
The tools this system runs on, grouped by the role they play.
Frontend
- Next.js
- React
- TypeScript
- Tailwind CSS
Backend
- Node.js
- REST API
- Background workers
AI
- OpenAI
- Embeddings
- Structured output
Data
- PostgreSQL
- Row-level security
- Full-text search
Integrations
- Twilio
- Email sync
- Webhooks
Infrastructure
- Vercel
- Docker
- GitHub Actions
Architecture
How It Works
The path a single request takes through the system, end to end.
SYNC01
Activity capture
Calls, email and web events flow into one timeline per account.
RESOLVE02
Identity resolution
Embedding-based matching links records and flags likely duplicates.
SUMMARISE03
Account narrative
A rolling summary is regenerated as significant activity lands.
SUGGEST04
Next action
Recommendations surface with the underlying signals attached.
REPORT05
Live reporting
Pipeline and performance views read from the same source of truth.
Outcome
Results
What changed for the business after launch.
3 → 1
Systems consolidated
One platform replaced a CRM, a shared inbox and a spreadsheet.
94%
Duplicate records resolved
Detected and merged during the migration and on an ongoing basis.
2 days
New-starter ramp time
Down from roughly two weeks of shadowing.
Live
Reporting cadence
Monthly manual reconciliation removed entirely.
Interface
A Closer Look
Screens from the delivered system.
Account view with rolling summary and suggested next action. 
Per-tenant pipeline configuration without code changes. 
Duplicate review queue with confidence and matching signals.
More work
Related Projects
Other systems built on similar foundations.
AI / SaaS
AI Customer Support Platform
A support platform that answers repeat questions from verified sources, drafts replies for agents, and escalates anything it cannot ground in documentation.
- Next.js
- Python
- Claude
- PostgreSQL
- +1
View Case Study
Next step
Want a System Like This?
Tell us what your business needs to automate and we'll map out a practical build.
hello@devrox.comWe reply to every enquiry within one business day.