The customer appeared twice. Orders sat in one place; shipping history sat somewhere else. In Amtex Systems’ account of a manufacturing project, two enterprise resource planning systems were feeding Salesforce, creating duplicate accounts and an incomplete view of the businesses buying steel. The sales team had software. What it needed was a customer record everyone could use.
- Amtex connects enterprise software, data and the people operating them.
- Its work spans analytics, cloud migration, engineering and staffing.
- The useful lesson: settle the meaning of the data before automating decisions.
It is a modest problem to put at the beginning of a company story. Nobody announces duplicate-record cleanup with a smoke machine. Yet it explains Amtex better than a catalogue of fashionable technologies. Much of enterprise computing happens in the awkward distance between a system being installed and a business being able to depend on it.
A customer before a company
Sunny Pokala knew that distance from banking. A 2007 Financial Express interview recounts his move from studying computer science to working in JPMorgan’s equity derivatives group. In 1997, he put $60,000 of savings into Amtex. When he discussed his plans with JPMorgan, the bank asked him to build something for it first.
He subsequently took the technology to other financial institutions. The interesting turn was repetition: a solution could travel across businesses. That experience helped turn a venture built around a technology problem into a technical consulting company. The early customer supplied both a problem and a reason to keep building.
“My company implemented the same solution across different businesses.”Sunny Pokala, in a 2007 interview
The repair beneath the dashboard
Return to those steel-company records. Amtex describes using Heroku to connect Salesforce with PostgreSQL, then manipulating records in bulk. Python scripts helped detect possible matches. Related shipments and orders were brought under parent accounts, producing a consolidated customer view.
The detail worth pausing over is the provision for manually merging and unmerging accounts. Automated matching needs a correction mechanism. A system that combines two genuinely different customers can create another problem while appearing to remove one. Amtex’s case includes permissions controlling who can handle the records, too.
The failure here was visible before any ambitious AI project: people could not reliably assemble the account history. The lesson is useful beyond Salesforce. Define what belongs together, preserve the relationships, and give someone authority to correct the result.
Old reports carry institutional memory
Amtex’s analytics offering addresses a related difficulty. A legacy report may contain a business rule that nobody wrote down elsewhere. Moving the report to a modern platform requires preserving what the calculation means. A prettier chart is little comfort when finance no longer trusts the total.
The company identifies WebFOCUS migration as a particular strength and lists Power BI, Microsoft Fabric, Tableau and Sigma among its target environments. Its published sequence starts by cataloguing reports, data sources, rules and users. Next comes rationalization: deciding what to keep, consolidate or retire. Migration includes checking outputs against the old system.
For a reader planning similar work, the copyable move is that second step. A migration is an opportunity to stop maintaining reports nobody needs. That decision requires business participation. Engineers can reproduce a calculation; the organization must decide whether it still answers a useful question.
Who buys the work behind the software?
Amtex serves enterprise and public-sector customers across finance, healthcare, manufacturing, retail and media. Its business combines consulting, custom development, implementation, staffing and continuing support. A customer might need an application built, workloads moved to the cloud, or specialists hired to keep a project moving.
Cloud services include readiness assessments, architecture audits, workload migration, validation, monitoring and disaster recovery. Product engineering extends into embedded software, verification and product lifecycle work. The delivery relationship can continue after launch through support and maintenance.
That puts Amtex among systems integrators, specialist consultancies and technology staffing providers. Its distinguishing proposition is the combination: it can work on the data, the application and the staffing problem around them. Its Salesforce practice also offers managed application support. The buyer is paying for implementation knowledge and ongoing capacity as well as code.
Care, tests and a separate crisis room
The portfolio makes the breadth tangible. Securra Health is a connected-care and telehealth platform with scheduling, video consultations and patient monitoring. Panther addresses repeated software testing, including functional checks, data validation and visual interface testing. These serve different users, but both turn recurring work into a software-supported process.
Public safety supplies another example. Amtex reported developing Kavalan, Tamil Nadu Police’s mobile safety application, in 2019, alongside a police master control room. That year it also announced receiving the CII CONNECT award for technology application.
In March 2023, Amtex announced a partnership to resell and implement CYGNVS in India and the Middle East. The platform provides cyber-response coordination outside normal internal communications. Teams can prepare playbooks and organize responders in a separate environment during an incident. The practical question is compelling: can the emergency team still coordinate when everyday systems become part of the emergency?
The meeting that comes before the model
Amtex’s current data and AI practice joins governance, data quality, engineering and data science. It describes building agents on governed enterprise data, with ownership, permissions, monitoring and output validation. The dependency matters: automation inherits the definitions and access rules beneath it.
My reading of these offerings is that the difficult meeting often comes before the technical demonstration. People need to agree what a customer is, which figures count, and who may change them. If an organization cannot make those decisions, a new platform leaves the dispute intact. Amtex’s most instructive work begins at that ordinary point: getting the records, and the people using them, to agree.