Optimised cloud infrastructure and shipped WhatsApp AI agents for a B2B logistics platform
Cloud and AI engagement for a B2B courier and logistics operator, covering cloud infrastructure optimisation and two AI agents on WhatsApp: a voice agent for customer enquiries and a B2B ordering agent for repeat business customers.
Key Takeaways
Quick Couriers operates a B2B logistics platform where customer enquiries and repeat ordering happen at high volume across WhatsApp. Unico Connect optimised the cloud infrastructure underneath the platform and built two AI agents on WhatsApp: a voice agent that handles customer enquiries and a B2B ordering agent that lets repeat customers place orders conversationally.
The combination lowers cost-to-serve on routine enquiries, captures B2B orders in the channel customers already use, and keeps the platform running on a tuned cloud footprint.
The Challenge
B2B logistics customers behave differently from consumers. They order frequently — often daily — with predictable patterns and the same operational questions: where is my package, when is the pickup window, what is the billing status. And in India they do this through WhatsApp, because that is where business happens.
Routine interactions were flowing through human agents on a contact-centre model, which carried two costs: higher cost-to-serve on each routine enquiry, and a slower customer experience as questions queued. Neither was sustainable as the platform scaled.
Quick Couriers came with a two-track ambition: optimise the cloud platform underneath the business, and build the WhatsApp AI agents that would shift routine load off human agents and into a channel customers already used.
Our Approach
We ran two parallel tracks rather than sequencing them — the cloud optimisation and the AI agents were both ready and neither blocked the other.
Key decisions:
Native to WhatsApp, integrated with the platform
Agents operate inside WhatsApp and connect to the platform’s real-time order and operations data, rather than redirecting customers to a separate interface.
Voice-first for this market
B2B customers here communicate through voice notes, so the voice agent handles speak-to-text and voice messages, and the ordering agent surfaces each returning customer’s typical patterns.
Targeted cloud tuning
We started from the workloads driving the largest AWS spend — instances, storage and operational settings — without changing anything affecting reliability without an explicit trade-off discussion.
The solution we built
A cost-optimised AWS footprint under the existing logistics platform (Laravel, Postgres, Bootstrap), plus two WhatsApp AI agents delivered through the WhatsApp Business API and integrated with live platform data.
WhatsApp voice agent
Transcribes and interprets voice enquiries, looks up real-time data and answers status, pickup-window and billing questions — escalating to a human with context when needed.
B2B ordering agent
Recognises returning customers, surfaces their typical order patterns and confirms conversationally before pushing the order through the platform.
Cloud optimisation
Right-sized instances, tuned storage and operational settings reduce the AWS cost footprint without affecting reliability.
Platform integration
Both agents reflect real-time order and operations state rather than running as standalone bots.
Tech stack
Outcomes & metrics
2 agents
WhatsApp voice + B2B ordering agents live
AWS
Cloud infrastructure optimised for cost-performance
Lower
Cost-to-serve on routine enquiries
