Platforms we have shipped, with AI in the loop.

The challenge

The hospital needed one system for OPD, IPD, pharmacy, lab, radiology, theatres, finance and staff, built to standard Indian hospital practice and ready to serve several hospitals from day one.

How we approached it

• ▸Multi-tenant architecture: every record carries its hospital, so a group runs on one platform.
• ▸Short sprints, each shipping a working module the hospital could try the same day.
• ▸Interoperability first: ABDM HIP/HIU with FHIR bundles, HL7 v2 analyser interfaces over MLLP, DICOM worklist for imaging.

AI in the build

• ✦AI clinical assistant and decision-support rules
• ✦Voice dictation with “Structure with AI” for clinical notes
• ✦“Ask the record”: retrieval-augmented answers over a patient’s own history
• ✦Document AI that reads outside prescriptions and reports into ward orders
• ✦AI first drafts for radiology and operative notes, always signed by a clinician

Results

• [ADD RESULT — e.g. registration time, report turnaround, go-live date]

The challenge

A business-critical Arabic and English site on an ageing CMS version, a hard eight-week window, and limited access to the client’s cloud environment.

How we approached it

• ▸Parallel build on our own environment, with the client’s Azure environment needed only before UAT.
• ▸Content and code migrated with Kentico Copilot, reviewed by engineers.
• ▸Lean team of three: lead developer, developer and shared PM/QA.

AI in the build

• ✦Kentico Copilot for code and content migration
• ✦AI-assisted QA of migrated pages

Results

• [ADD RESULT — e.g. launch date, Lighthouse score, editor time saved]

The challenge

A plugin-heavy WordPress site that was slow to change, with Arabic and right-to-left support as a must-have.

How we approached it

• ▸Component-based React front end with a .NET Core API and SQL Server.
• ▸Arabic RTL designed in, not bolted on.
• ▸A developer-facing API playground for the product.

AI in the build

• ✦AI-assisted content migration and translation review

Results

• [ADD RESULT — e.g. page speed, conversion rate, time to publish]

The challenge

Regulated invoice data needed to live in-country with clean separation between environments.

How we approached it

• ▸Dedicated subscription with consistent naming
• ▸All resources in the UAE region
• ▸Separate Dev and Live environments

AI in the build

• ✦[ADD — e.g. AI invoice extraction or anomaly checks, if used]

Results

• [ADD RESULT]

The challenge

Regulated organisations want AI but cannot send data to public endpoints.

How we approached it

• ▸On-premise and private-cloud deployment patterns
• ▸Sector-specific pilot scopes
• ▸PDPL compliance checklist for every pilot

AI in the build

• ✦Private LLM assistants
• ✦Document understanding
• ✦Retrieval over internal knowledge

Results

• [ADD RESULT]