AI Engineer

Capgemini; Hyderabad; On-site; INR 2,800,000 - 4,000,000; Full time

Benefits offered by this job

Competitive salary and performance‑drI

Comprehensive benefits package

Career development and training

Flexible work arrangements (remote and

Dynamic and inclusive work culture

Private Health Insurance

Retirement Benefits

Paid Time Off

Training & Development

Benefits vary by level

Job summary

Capgemini seeks an AI Engineer to design, build, and ship production AI systems—not prototypes. You’ll own the path from idea to a reliable, load-tested production system with measurable performance and guardrails.

We want engineers who have shipped agentic or LLM-powered systems in production and can justify architectural choices. The role requires strong software engineering, Python/TypeScript, and hands-on AI tooling expertise, in a collaborative, high-scrutiny environment.

Qualifications

• Design and build production-grade AI/agentic systems from architecture to deployment.
• Own full lifecycle of at least one non-trivial AI capability: problem framing, evaluation, prompt/context engineering, orchestration, deployment, and post-launch tuning.
• Build deterministic guardrails around probabilistic components—retries, validation, fallback paths, human-in-the-loop checkpoints where confidence is low.
• Design evaluation harnesses and offline/online eval pipelines that catch regressions, not vanity metrics.
• Make real architectural tradeoffs on latency, cost, and reliability—token economics upfront.
• Integrate AI systems into existing production infrastructure (APIs, data pipelines, orchestration layers) without breaking existing work.
• Push back on bad ideas with technical reasoning, not opinions.

Responsibilities

• Design and build production-grade AI/agentic systems—from architecture through deployment, monitoring, and iteration.
• Own the full lifecycle of at least one non-trivial AI capability: problem framing, evaluation strategy, prompt/context engineering, orchestration, deployment, and post-launch tuning.
• Build deterministic guardrails around probabilistic components—retries, validation, fallback paths, human-in-the-loop checkpoints where confidence is low.
• Design evaluation harnesses and offline/online eval pipelines that actually catch regressions, not vanity metrics.
• Make real architectural tradeoffs on latency, cost, and reliability — token economics is a design constraint you think about upfront.
• Integrate AI systems into existing production infrastructure (APIs, data pipelines, orchestration layers) without breaking what already works.
• Push back on bad ideas — including ours — with technical reasoning, not opinions.

Skills

LLM agentic systems

Python

TypeScript

APIs integration

LangGraph

CrewAI

Vector stores

RAG pipelines

Prompt engineering

Evaluation methodology

CI/CD

Observability

Cost latency tradeoffs

Tools

Kubernetes/OpenShift

AWS/Azure/GCP

Portkey

LangGraph

CrewAI

Strands

Vector stores

Job description

Function: AI Engineering / Applied AI Delivery

About The Role

We're hiring an AI Engineer to design, build, and ship production AI systems — not prototypes, not notebooks. This is a builder's role: you'll own the path from \"we think an LLM/agent could solve this\" to a system running reliably in production, under real load, with real failure modes.

We are being deliberately selective here. This role is not for someone who has \"used ChatGPT a lot\" or built a weekend RAG demo. We want people who have shipped agentic or LLM-powered systems that other engineers depend on, who understand why those systems break, and who can hold their own in a room full of skeptical senior engineers. If that's not you yet, this probably isn't the right role yet either — and that's fine.

Requirements

What You'll Own

• Design and build production-grade AI/agentic systems — from architecture through deployment, monitoring, and iteration — not just model calls wrapped in a script


• Own the full lifecycle of at least one non-trivial AI capability: problem framing, evaluation strategy, prompt/context engineering, orchestration, deployment, and post-launch tuning


• Build deterministic guardrails around probabilistic components — retries, validation, fallback paths, human-in-the-loop checkpoints where confidence is low


• Design evaluation harnesses and offline/online eval pipelines that actually catch regressions, not vanity metrics


• Make real architectural tradeoffs on latency, cost, and reliability — token economics is a design constraint you think about upfront, not an afterthought


• Integrate AI systems into existing production infrastructure (APIs, data pipelines, orchestration layers) without breaking what already works


• Push back on bad ideas — including ours — with technical reasoning, not opinions

Who You Are

• 4+ years of strong software engineering experience, with at least 1-2 years hands-on building and shipping LLM/agentic systems in production (not just experimentation)


• Fluent in Python and/or TypeScript, with the engineering discipline to write systems that survive contact with real users


• Real, hands-on depth with modern AI tooling — LLM APIs (Anthropic, OpenAI, etc.), agent frameworks (LangGraph, CrewAI, Strands, or equivalent), vector stores/RAG pipelines, and prompt/context engineering — and a clear, opinionated point of view on where each of these breaks down


• Strong grasp of evaluation methodology — you know the difference between a model that looks good in a demo and one that's actually reliable


• Comfortable with orchestration and systems fundamentals: APIs, event-driven design, queuing, observability, CI/CD


• Able to reason clearly about cost, latency, and failure modes at design time, not just after something breaks in production


• Sharp communicator — can explain a technical tradeoff to both an engineer and a non-technical stakeholder without dumbing it down or overcomplicating it


• Genuinely curious and self-directed — this space moves weekly, and we need someone who tracks it because they want to, not because it's a KPI

Nice to Have

• Experience with cloud-native deployment (Kubernetes/OpenShift), and cloud platforms (AWS/Azure/GCP)


• Exposure to AI gateways / model routing layers (Portkey or equivalent)


• Experience with structured spec-driven or agent-first SDLC platforms


• Contributions to open-source AI tooling, published technical writing, or a portfolio of shipped AI products you can speak to in depth


• Experience in regulated industries (finance, insurance, healthcare) where reliability and auditability are non-negotiable

Working Style

• Onsite Pune or Hyderabad, India.


• Direct, low-ceremony communication. We'd rather hear \"this approach is wrong and here's why\" in week one than a polished status update

Benefits

This position comes with competitive compensation and benefits package:

• Competitive salary and performance-based bonuses


• Comprehensive benefits package


• Career development and training opportunities


• Flexible work arrangements (remote and/or office-based)


• Dynamic and inclusive work culture within a globally known group


• Private Health Insurance


• Retirement Benefits


• Paid Time Off


• Training & Development


• *Note: Benefits differ based on employee level

About Capgemini

Capgemini is a global leader in partnering with companies to transform and manage their business by harnessing the power of technology. The Group is guided everyday by its purpose of unleashing human energy through technology for an inclusive and sustainable future. It is a responsible and diverse organization of over 420,000 team members in more than 50 countries. With its strong 55-year heritage and deep industry expertise, Capgemini is trusted by its clients to address the entire breadth of their business needs, from strategy and design to operations, fueled by the fast evolving and innovative world of cloud, data, AI, connectivity, software, digital engineering and platforms. The Group €22.5 billion in revenues in 2025.