Senior Artificial Intelligence Engineer

YottaFlex AI Technologies Pvt. Ltd; Hyderabad; On-site; INR 1,400,000 - 2,100,000; Full time

Job summary

YottaFlex AI Technologies Pvt. Ltd in Hyderabad, India, seeks a genuinely full-stack, AI-first engineer to ship production code daily across the full stack and the AI layer. You will prototype fast, then harden it into a multi-tenant, observable service with enterprise-grade quality and security.

You will work with Python/TypeScript, React, REST/GraphQL, AWS Bedrock, Lambda, ECS/EKS, and CI/CD pipelines. This role emphasizes real ownership, fast delivery, and rigorous testing without slide decks.

Qualifications

• Production experience with frontier models in live deployments.
• Experience designing multi-tenant SaaS with per-tenant isolation.
• Familiarity with retrieval and RAG pipelines and memory design.
• Knowledge of LLM risk categories and observability tooling such as LangSmith/Langfuse.

Responsibilities

• Ship production full-stack code daily across the AI layer and services.
• Prototype quickly, then harden into scalable, observable, enterprise-grade systems.
• Own deployment, monitoring, incident response and postmortems that drive improvements.

Skills

Python

TypeScript

React

REST

GraphQL

gRPC

AWS

Bedrock

LLM observability

Vector search

Tools

Docker

Terraform

LangSmith

Langfuse

CI/CD

Job description

About the role

This is a coding role. You will ship production code every day across the full stack and the AI layer. You take an ambiguous problem, prototype a working answer fast, then harden it into a multi-tenant, observable, enterprise-grade service. This role does not drift into slides.

We need someone who is genuinely full stack and genuinely AI-first. Not a backend engineer who has called an LLM API once, and not a researcher who has never carried a pager.

Core tech stack

Languages: Python and TypeScript, fluent in both, not passable in one. Node where it fits.

Frontend: React (or equivalent modern framework), streaming and async UX for AI interfaces.

Backend and APIs: REST, ideally GraphQL or gRPC. Async and event-driven patterns, caching, queueing.

Data: Relational modelling, vector search and retrieval, ETL and pipeline basics.

Cloud: AWS at depth, including Bedrock, Lambda, ECS or EKS, and managed data services. Working exposure to a second cloud (Azure or GCP).

DevOps: Docker and container orchestration, CI/CD, infrastructure as code with Terraform, CDK, or similar.

AI engineering you must have

Deep production experience with frontier models (Claude, GPT, Gemini, or comparable), running live, not in notebooks. Specifically:

• Agentic patterns, orchestration, tool and function calling
• Retrieval and RAG pipelines, context and memory design
• Structured output, streaming, and fallback behaviour when the model gets it wrong
• Deterministic guardrails around non-deterministic systems
• LLM observability and tracing with LangSmith, Langfuse, or equivalent
• Evaluation as engineering practice: eval sets, release gates, regression suites
• Cost and latency engineering: token budgeting, prompt caching, batching, model routing
• Fluency with AI coding tools in your own workflow, with opinions on where they help and where they hurt

Also required

Multi-tenant SaaS: Real tenant isolation, per-tenant config, usage metering, tenant-scoped auth, cross-tenant migrations.

Testing: Unit, integration, and contract tests as how you move fast, not a tax on speed.

Security and compliance: Prompt injection, data leakage, and PII handling by design. Familiarity with the OWASP LLM risk categories, SOC 2, audit trails, data residency, and least-privilege access.

Ownership: Monitoring, alerting, incident response, and postmortems that change something.

Nice to have

MCP and tool servers for agentic systems, developer tooling or SDLC

platform experience, fine-tuning or serving open-weight models, B2B SaaS into regulated buyers (financial services or government), team lead or mentoring experience, open source or public technical writing.