Research & development · RD 001–006
Technology research and product development
We study a problem before we build for it. Not because research sounds impressive, but because most software fails for reasons that were visible before anyone wrote a line of it.
The process
How an idea becomes a product.
Research
Define the problem and study how it is currently solved, including the workarounds people have already invented.
Prototype
Build something rough and fast to test the idea against reality rather than against a document.
Build
Turn a validated prototype into maintained software, with documentation and a handover path.
Operate
Run it under real usage and real failure conditions, and keep developing it. Software is not finished at launch.
Research areas
Six areas, each with something shipped behind it.
RD-001 · Artificial intelligence
Applied AI in customer operations — response, routing, qualification and support. The unglamorous applications, where the return is measurable.
RD-002 · Business automation
Removing repetitive, rule-based work from operational processes, and designing the failure path for when automation gets something wrong.
RD-003 · CRM systems
Systems built around how a business actually works rather than how software expects it to, particularly for long and non-linear sales cycles.
RD-004 · Digital commerce
Catalogue data, conversion and the technical layer underneath online retail, including the relationship between speed and revenue.
RD-005 · Growth systems
Acquisition treated as an engineering problem — attribution, capture, routing and follow-up as a connected system rather than as campaigns.
RD-006 · Digital experiences
Interface, performance and accessibility, and the measurable relationship between how fast something loads and whether it converts.
Two offices
Where the products get built.
Our product engineering team works from Jaipur. Research, prototyping and the software itself happen there, on a deliberately narrow focus — one product taken to a stable release before the next one starts.
Kolkata handles deployment, integration and the campaigns that run through those systems. So the people who write the software and the people who watch it perform under real traffic are in the same company, working against the same backlog.
That is an unusual arrangement, and it is the reason we can say we build the system and then run it. Most companies can only do one of the two.
Current work
What we are working on now.
[CONTENT TO VERIFY — populate with current research, and commit to quarterly updates.]
A section like this is only worth having if it is maintained. An unchanged "what we are working on" page is a clearer signal of an inattentive company than no page at all, so this should be updated quarterly or removed.
Have a technical problem worth studying?
The research stage is small, bounded, and about a third of the time concludes that an existing product already solves it.