Our work

Projects we've shipped & built

From AI calling agents to cloud security tools — real projects built with production-grade engineering.

How it works

Built using LangChain and OpenAI APIs with RAG pipelines that give the agent real-time access to product knowledge, pricing, and CRM data. Integrated with telephony systems via Node.js CRUD APIs for real-time call execution. Implemented TTS/STT conversational flows for seamless voice interaction.

Key features

Natural voice conversations with customers
RAG-powered product & pricing knowledge
Real-time CRM integration via Node.js APIs
TTS/STT pipeline for end-to-end voice flow

How it works

Backend CRUD APIs built in Node.js handle all expense data. OpenAI APIs intelligently categorise expenses from natural language input and receipt photos. A RAG pipeline built over the user's financial history powers a personalised AI insights feature that answers questions like "Where am I overspending?" and "Can I afford this?"

Key features

Automated expense logging via email parsing
Voice-based expense input with AI categorisation
Photo receipt scanning and data extraction
RAG pipeline for personalised financial insights

How it works

Prototyped using LangChain + OpenAI for natural language understanding and order-taking. FastAPI handles LLM communication and backend orchestration. A RAG workflow over the live menu, pricing, and active promotions ensures the agent always responds with accurate, context-aware information.

Key features

Real-time order-taking via natural conversation
RAG-grounded menu, pricing & promotions
Handles order modifications and clarifications
FastAPI backend for low-latency LLM calls

How it works

Built the full backend on Firebase with a schema designed to support advanced health analytics as the app scales. Integrated OpenAI API to power a Health Agent feature that generates personalised recommendations based on user health data and patterns.

Key features

Future-proof Firebase schema for health analytics
AI-driven personalised health recommendations
OpenAI-powered Health Agent feature
Privacy-first data architecture

How it works

Built using IAM and Azure AD role analysis to map actual permissions against least-privilege best practices. Implemented automated test suites that scan for common misconfigurations — including publicly accessible Azure SQL databases, over-permissioned IAM roles, and exposed storage buckets — and generate remediation reports.

Key features

Automated misconfiguration detection on AWS & Azure
IAM/AD role analysis against least-privilege policy
Prevents public exposure of Azure SQL and storage
Generates actionable remediation reports

Have a project in mind?

Tell us what you're building. We'll scope it, price it, and get it shipped.

Start a Project →View Case Studies
Chat on WhatsApp