Who gets hired, and why
Java backend and full-stack
Full-stack for the build
.NET
GenAI
Data
Salesforce and platforms
Generalists
Test
Embedded and systems
JavaScript and Node full-stack
Mobile
Frontend
Cloud and DevOps
Go and mainframe
Java backend and full-stack engineers. The build is usually a Java system, and the client's engineers who will take it over work in Java too, so these firms hire Java engineers in depth. Most of the openings are pure backend work in Spring Boot; the rest are full-stack with React or Angular on top. The postings go far beyond the basics of the framework. Many require Spring Security, Spring Cloud and JUnit, with REST Assured for testing the services, and most require Kubernetes, a cloud and a deployment pipeline, because the team ships what it builds. Java is the one role every firm here hires for, from EPAM to the studios, like InRhythm and Fluxon, and it is the signature of the build shop.
Full-stack and backend engineers hired for the build, not the language. A good number of openings name the product to be built rather than the language. These engineers work in React and TypeScript with Python or Java behind, and the postings require Terraform and Argo CD, the tools that put a build into a client's cloud. They appear here at about twice their weight across the services world, and they are almost all at EPAM.
.NET engineers. Banks, insurers and retailers that run on the Microsoft stack hand their rebuilds to these firms, so .NET engineers appear at about their usual weight. Nearly all the openings are full-stack, with React or Angular on top of C# and ASP.NET Core, and Azure beside them. A notable share of these postings also require experience of building features with large language models, more than for any other role here except GenAI engineers.
GenAI engineers. A build firm that builds with AI is what every client now wants, and these firms hire GenAI engineers at about the services rate. They build applications on top of large language models, in Python with LangChain and the OpenAI and cloud model services, and most of the postings also require backend and data work, since the model sits on the client's data. The work is spread across EPAM and the newest studios alike, and the AI share of hiring at EPAM has risen through 2026.
Data engineers. When a client's build is a data platform, these firms supply the engineers who make it. The openings require Python above all, with Spark, Hive and Hadoop, AWS or Azure, and data warehousing. They are at about their usual weight, nearly all at EPAM, and they are among the hardest jobs here to fill.
Salesforce engineers and the other packaged platforms. A part of what EPAM builds is on platforms a client has bought rather than written, and that gives this Company Tier a Salesforce role at several times its weight across the services world. The postings require Apex, Lightning Web Components, SOQL and the Sales and Service Clouds. Beside Salesforce sit Workday and Oracle developers, enterprise application developers who connect and extend a client's business systems in Java, Spring and SQL without a named platform, and a few SAP engineers on ABAP and HANA. Nearly all of this platform work is at EPAM; the rest of the Company Tier posts almost none of it, so a reader with a Salesforce or SAP background should not expect it of a build shop in general.
Generalist software engineers. Some openings are for engineers to join a client's product team outright, in whatever the client uses, typically Python, Java or TypeScript with Kafka and Azure DevOps. They appear at above their usual weight and are concentrated at Luxoft, where a handful of them are trading-system engineers who must know the FIX protocol, the message standard that trading systems use to talk to one another.
Test engineers. Testing is a smaller part of a build than of a run, so test engineers appear well below their weight across the services world. What there is leans modern. The postings require Playwright and Cypress as often as Selenium, with Java, Jenkins and GitHub Actions, and some ask for performance testing. The largest group of these openings is at Luxoft, not at EPAM, which posts hardly any.
Embedded and systems engineers. The firms that build for carmakers, device makers and network companies, like Luxoft, GlobalLogic and Ascendion, hire embedded firmware engineers and C and C++ engineers on Linux. The firmware postings require CAN bus, embedded Linux, device drivers and a real-time operating system, with AUTOSAR for the automotive work; the systems postings require TCP/IP and the Linux kernel. Together they are a modest role here, below the services rate, but they are where the narrative's carmaker and device work shows up in the data.
JavaScript and Node full-stack engineers. A small role, at about its usual weight, hired for builds in Node.js, React and TypeScript with GraphQL, at EPAM and at mid-sized firms, like Nagarro and Luxoft, and studios.
Mobile engineers. Fewer than across the services world, and concentrated at GlobalLogic, where the Android postings require Bluetooth, Wi-Fi and the Android platform build itself. This is device software more than consumer apps.
Frontend engineers. A small role, at about its usual weight, at studios and mid-sized firms rather than EPAM. The postings require React above all, with Cypress for testing and attention to performance.
Cloud and DevOps engineers. Strikingly few, at a small fraction of their weight across the services world. A build firm hands over what it builds and the client runs it, so the cloud work is written into the engineering roles (Kubernetes, Terraform, pipelines) rather than hired separately. The few openings require Python, Kubernetes, AWS and Terraform.
Go engineers and mainframe engineers. Two small roles worth naming. Go engineers build cloud services in Go with gRPC and Kubernetes. Mainframe engineers, in COBOL, JCL and IMS, appear mostly at EPAM, for banks' and insurers' older systems.
Rare here. Python backend engineers, ML engineers, security software engineers, ServiceNow engineers and Power Platform developers barely appear.
Where in the Company Tier the work is
Hiring here is concentrated in EPAM to a degree found in few other units. EPAM posts three in five of all openings, and its mix sets the Company Tier's. It is Java first and by a distance, then data engineering, Salesforce and the other packaged platforms, .NET, GenAI and full-stack work, with almost no testing, cloud or embedded work.
The rest of the Company Tier reads differently. Java still leads, though by much less, and testing is the second role, followed by .NET, generalist software engineers, embedded and systems engineers, cloud and DevOps, mobile and frontend. Much of the embedded, automotive and device work comes from two firms, Luxoft, which builds for carmakers and banks, and GlobalLogic, which builds device and mobile software, and Luxoft also posts much of the testing. The studios post small numbers across Java, frontend, GenAI and cloud.
Bengaluru has the most work, but less dominantly than elsewhere in the services world. EPAM hires more in Pune, Chennai and Hyderabad than in Bengaluru, which lifts those three cities to a near-equal second tier. The rest of the Company Tier is Bengaluru first by a wide margin, with Pune and Delhi NCR behind.