Careers - Xpergia

Vansh Seth

Intern

My time as an intern at Xpergia has been one of the most formative chapters of my journey as an engineer so far. From day one, I was treated less like an intern fetching tickets from the periphery and more like a contributing member of the engineering team, trusted with real production systems, real customers, and the real consequences that come with both. The work was challenging, the pace was fast, and the bar for quality was high, but the environment around it was uniquely supportive: code reviews that taught me something on every PR, mentors who explained the "why" behind architectural decisions, and a culture that consistently rewarded curiosity over caution. The bulk of my work centered on a sports analytics and AI-driven prediction platform, where I owned end-to-end feature delivery across a Python backend that powers a large catalog of LLM-driven prediction tools. Over the course of the internship, I designed and shipped well over thirty distinct prediction markets and their supporting API. Each of these was more than just a function – it required carefully shaping the prompt context fed to the LLM, defining the YAML tool schemas, wiring up the tool executor, building the corresponding REST endpoints and controllers, and writing the head-to-head data fetching logic that backed the predictions. Along the way I picked up a strong intuition for how to make LLM tool-use reliable in production: pre-formatting player lists, limiting context to top-N candidates, removing noisy thresholds from prompts, and ensuring every market produced consistent, explainable outputs. Beyond shipping features, I contributed substantially to the platform's reliability and engineering hygiene. I authored and ran end-to-end production load tests using JMeter, designed for sustaining a target of 50,000 concurrent users, and helped stand up the Datadog observability stack that supported them. I built out a load-generation VM, wired the JMeter harness into the production APIM, and contributed to the RTO test reports and data-copying reports that gave the team confidence in our disaster-recovery posture. On the infrastructure side, I worked on creating a hardened production environment on Azure AKS – coordinating across dedicated PostgreSQL, Redis, Event Hub, API Management, and Azure Front Door with WAF – and tightened monitoring and alerting around it. I also authored the GitHub Actions workflows that automated deployments to production AKS across multiple repositories, which significantly shortened our release cycle and removed an entire class of manual-deployment errors. A meaningful portion of my time was also spent in the data ingestion pipeline, where I diagnosed and fixed a series of subtle but high-impact bugs: normalizing inconsistent stat shapes coming from upstream providers, converting match start and end timestamps to proper `timestamptz` so timezone behavior was correct end-to-end, repairing broken backend entrypoint paths, and improving exception logging in the concurrent processing path so silent failures stopped being silent. These were the kinds of issues that don't make for flashy demos but quietly determine whether a data platform is trustworthy, and learning to hunt them down – and to write the migrations and tests that prevent them from recurring – was one of the most valuable parts of the experience. I additionally contributed to the live-match prediction service, building tools around in-play card risk and player foul markets and helping introduce baseline prompt scaffolding for the live LLM client. What I'll carry forward from Xpergia goes well beyond the technical surface area, although that surface area was wide: Python, FastAPI-style backends, LLM tool calling and prompt engineering, Azure AKS and APIM, PostgreSQL, Redis, Event Hub, Front Door with WAF, Datadog, JMeter, GitHub Actions, Ruff/Flake8/Black tooling, and disciplined PR-driven development. More importantly, I learned how a serious engineering team thinks: how to break a vague product ask into a clean, well-scoped PR, how to write code that a reviewer can actually review, how to push back respectfully when a design feels wrong, how to reason about cost, scale, and blast-radius before merging, and how to keep shipping under pressure without cutting the corners that matter. Xpergia gave me ownership early, trusted me to grow into it, and surrounded me with engineers who genuinely cared about making me better. I'm leaving the internship a markedly stronger developer than I came in, and deeply grateful to the team for the trust, the mentorship, and the room to do meaningful work.

Kaushik Kadium

Intern

My experience at Xpergia has been incredibly valuable for my professional growth. From my very first day, I was welcomed not just as an intern, but as a core contributor trusted with significant responsibilities. The culture here is exceptionally supportive; the management and engineering teams are highly responsive and consistently foster an environment where every voice is respected. Being entrusted with the end-to-end responsibility of a complete project as a fresher is a rarity in the industry, and it speaks volumes about Xpergia’s deep trust in its people. Coupled with a strong emphasis on professional work ethics and a genuinely healthy work-life balance, this environment has made my time here truly splendid. My primary focus was architecting and developing a highly secure, serverless enterprise dashboard designed for external plant managers. I took ownership of the frontend delivery, building a clean, intuitive user interface that integrated seamlessly with our Python-based API engine.Under the guidance of my managers, I developed a strong design philosophy rooted in empathy for the end-user. They consistently encouraged me to step into the customers' shoes, which taught me how to identify genuine user-experience friction points and engineer solutions that clearly elevated the dashboard's usability. This user-first mindset was critical in transforming complex, high-stakes data into a clean, intuitive, and accessible interface. Beyond the codebase, this project served as a masterclass in professional software engineering practices. I gained extensive hands-on experience with enterprise version control, learning how to effectively manage Git repositories, structure clean commits, and navigate the rigorous pull request (PR) and code review processes. While delivering this seamless UI, I also contributed to our cloud infrastructure, integrating tools like AWS Cognito to establish a secure, "Zero-PII" authentication layer. Perhaps the most defining aspect of my internship was the creative autonomy I was afforded. At Xpergia, having a good idea means having the freedom to actually build it. Management actively encouraged me to express my technical proposals, which were not just heard, but fully supported and directly integrated into live projects. Seeing my own architectural and design ideas trusted and deployed in production is a true testament to how deeply Xpergia values, respects, and empowers every member of its team.

Build AI with people who care

Open roles across our team

Senior ML Engineer

AI Solutions Architect

Implementation Consultant

Data Engineer

Product Manager - AI

Four principles guiding every partnership

People First, Always

Growth Mindset

Collaboration Over Competition

Quality With Purpose

Benefits for your best work

Remote-First Culture

Learning & Development

Competitive Compensation

Health & Wellbeing

High-Impact Work

Collaborative Team

Hear from our recent interns

Vansh Seth

Intern

Kaushik Kadium

Intern

Don't See the Right Fit?