The trouble at the pharmacy benefits manager was almost embarrassingly ordinary. Member history lived in disconnected systems. Outreach involved manual steps. Claims had coding errors. As membership grew, the work threatened to demand more people in roughly the same proportion. In Veltris’s account of the engagement, growth itself had become an administrative problem. The interesting question was where to intervene.
- Veltris builds software and connects the data underneath it.
- Healthcare, secure networks and industrial operations are its main hunting grounds.
- Its AI proposition depends on understanding the job being done.
Veltris built a care operations platform around a shared data foundation, automated outreach and agents assisting claims and coding. People retained review of exceptions. The company reports three times the member portal adoption and 30% fewer claim denials. Those are results from one company-published case, rather than promises for the next buyer. Still, they suggest something useful: the opportunity was hiding between tasks.
01 / Two firms, one longer to-do list
Veltris entered the world through a combination. In October 2023, Trive Capital and BayLink Capital announced the joining of Wavelabs Technologies and West Agile Labs. Wavelabs, founded in 2015, brought engineering R&D and AI capabilities. West Agile, founded in 2012, brought product architecture, design, application development and DevOps. The new business could work from embedded software inside hardware to applications in the cloud.
That history explains a small dating trap. Veltris’s formation belongs to 2023; the older engineering histories belong to its constituent firms. Mansoor Khan became the combined company’s CEO and Pramod Dabir its president. In September 2024, Hiral Chandrana, previously CEO of Mastek, took over as CEO and joined the board. This was an assembled engineering business with an acquisition strategy from the outset.
02 / The dental detail matters
The strategy sharpened in March 2025 with the acquisition of BPK Technologies. BPK specialized in custom software for dental service organizations, practice-management systems and health systems, alongside education and banking. Its capabilities included insurance verification, claims automation and data warehousing. For a company talking about AI, that is a wonderfully specific shopping list. Knowing how a dental practice gets paid is part of knowing what to automate.
Read as a sequence, the deals show a move from engineering breadth toward industry depth. Veltris’s current focus includes pharmacy benefits, dental and retail health, secure networking and manufacturing. Its customers include enterprises, mid-market businesses and private equity portfolios. These organizations often bring accumulated systems and operating habits. The engineering assignment is to make those parts cooperate while the business keeps running.
One industrial case makes that inheritance visible. An energy infrastructure services company had consolidated seven regional providers, leaving fragmented financial systems, slow invoicing and manual dispatch across more than 21 offices and a 1,400-person field workforce. Veltris describes connecting operational and asset data, automating billing and supporting field-service optimization. For a buyer growing through acquisitions, this is a recognizable assignment. Adding companies expands the organization before it creates a shared operating system. Connecting the records becomes a prerequisite for using them intelligently, and the invoice waiting to be sent gives the project an immediate business purpose.
- 01Connect the recordsBring fragmented operational data together.
- 02Automate a specific jobClaims, outage analysis or quality inspection.
- 03Keep judgment in the loopRoute exceptions to people; measure the result.
03 / A platform with a day job
In May 2026, Veltris launched VeltrisOne, its vertical AI orchestration platform. The phrase sounds abstract; the applications are fairly concrete. HelixFlow addresses healthcare administration and revenue-cycle work. FaultAnalyze examines network faults and outages. DefectQC uses computer vision for manufacturing inspection. The platform brings applications, agents and workflows together, built on Microsoft Azure AI and Databricks, with integration into existing enterprise systems and an on-premises deployment option.
The commercial proposition is reuse with context. A new customer should not have to commission every connector and workflow from scratch. Yet an insurance claim and a factory defect require different data, rules and consequences. Veltris is trying to package the repeatable work while keeping the industry knowledge. Its launch lists as-a-service, managed AI platform and enterprise transformation models. Bespoke engineering and ongoing product support remain part of the wider business.
There is also a security obligation attached to connecting more things. A December 2025 partnership with Vectra AI added network, identity and cloud threat detection to Veltris’s managed security offering. That fits the same operating picture: the work continues after software goes live. Access, monitoring, maintenance and response become part of the service a buyer needs to evaluate.
04 / Four years beats one demonstration
A networking case provides a useful counterweight to the AI language. Veltris describes an engagement lasting more than four years, with its team growing from three engineers to 45. The work included modernizing a React interface, migrating APIs, developing networking features and building Python-based test automation. This is the unglamorous continuity of product engineering: ship changes, preserve quality, fix problems, repeat.
“They bring both the technical depth and the delivery consistency that keeps our roadmap moving.”
Anonymous networking executive, in Veltris’s case study
That is where Veltris sits in the market: a digital engineering partner combining specialist delivery teams with reusable software. Buyers can also build internally or hire another engineering firm. Veltris’s stated distinction is its combination of industry workflows, data modernization and implementation. The competitive test is practical: can that combination shorten the work and improve the product in the buyer’s actual environment?
05 / Take the workflow map to procurement
The lesson a reader can copy is refreshingly modest. Pick one expensive recurring process. Map its handoffs. Establish how often it fails, how long it takes and what an exception costs. Then ask which data and decisions an automated system would need. That sequence gives a buyer something more useful than a demonstration: a way to decide whether the proposed improvement is worth paying for.
It also defines the conditions for a sensible engagement. If records cannot be accessed, the process has no accountable owner or nobody agrees on a baseline, a packaged AI application will have little to stand on. Human review needs a place in the design. Veltris’s most persuasive proposition starts with this ordinary discipline. Follow the rejected claim, the repeated entry, the troublesome handoff. There is a business hiding in the inconvenience.