AI Native Engineer

Talismatic; Bengaluru; Hybrid; INR 1,200,000 - 2,400,000; Full time

Job summary

Zycus is hiring AI-Native Engineers to embed AI in the product end-to-end, from UX to production. You will build agents, tool calling, RAG, and evals to ship production-grade AI features across invoices, POs, contracts and supplier management.

Role emphasizes owning the full stack, from frontend React/TypeScript interfaces to Python/Java services, data models, and cloud deployment, with strong emphasis on reliability and governance.

Qualifications

• Hands-on experience building with LLMs/GenAI in production.
• Experience with agents, tool calling, RAG and structured outputs.
• Ability to design end-to-end capabilities from UX to production.
• Strong CS fundamentals: distributed systems, API design, data modeling.
• Proficiency in Python, Java or TypeScript and modern APIs.
• Experience with React front-ends and cloud deployments.

Responsibilities

• Build AI-native product capabilities end-to-end within the S2P suite.
• Design agents and agentic workflows using tool calling and RAG.
• Balance deterministic logic with AI judgment and define boundaries.
• Develop LLM-powered features integrated with enterprise workflows.
• Contribute across UI, backend services and data models.
• Ensure reliability: evals, guardrails, observability and auditing.

Skills

AI-Native engineering

LLMs / GenAI

Agents and agentic workflows

Tool/function calling

RAG / context engineering

Evals / evaluation

Model routing

Python/Java/TypeScript

React / UI design

Education

Bachelor’s or Master’s in CS/Engineering

Tools

React

TypeScript

Python

Java

APIs / microservices

Databases and cloud

Job description

Zycus is building a next-generation Agentic AI Platform to autonomously execute enterprise Source-to-Pay (S2P) procurement workflows.

We are looking for AI-Native Engineers who build AI into the product itself (agents, RAG, tool calling, evals) and own the whole flow from user experience to production. For these engineers, AI is both their tool and their material: they use AI coding agents to multiply what they can build, and they ship software that has AI inside it.

This role requires someone who can take an ambiguous enterprise problem, rethink the experience using AI, decide what should be agentic vs deterministic, and independently design, build, evaluate, debug and productionize the solution.

Roles & Responsibilities

• Build AI-native product capabilities for Zycus’s S2P suite end-to-end: UI → AI/agent layer → services/APIs → data → production

• Design and implement agents and agentic workflows using tool/function calling, RAG/context engineering, structured outputs and MCP (or equivalent)

• Decide which logic should remain deterministic (rules, validations) and which needs AI judgement, and design the boundary between them

• Develop LLM-powered features integrated with enterprise workflows (invoices, POs, contracts, sourcing, supplier management)

• Contribute across the stack, including:

  • Frontend interfaces (React / TypeScript) for AI-driven workflows

  • Backend services and APIs (Python, Java or TypeScript)

  • Data models, integrations and cloud deployment

• Ensure production-grade AI reliability:

  • Evaluations (evals), accuracy measurement and regression testing

  • Handling hallucinations, guardrails, fallbacks and auditability

  • Observability, monitoring and debugging of AI systems

• Select and route between models, optimizing for quality, latency and token/model cost

• Use AI coding agents (Claude Code, Cursor, Codex, GitHub Copilot or similar) as a core part of how you design, build, test and debug, while fully understanding and owning the code they produce

• Rapidly evaluate and adopt new AI and engineering technologies

• Share patterns, evals and tooling that raise the team’s AI-native capability (mentoring expected at Senior level and above)

Skills, Experience & Education (Required)

1. Experience

• 3+ years of hands-on software engineering experience, hiring across levels:

  • AI-Native Software Engineer: 3–6 years

  • Senior AI-Native Engineer: 6–10 years

  • Staff / Lead AI-Native Engineer: 8–13 years

  • Principal AI-Native Engineer: 10–15+ years

• We value engineering depth, learning velocity and what you’ve built recently over years of experience

2. AI-Native Engineering (Core)

• Substantial hands-on experience building applications with LLMs/GenAI

• Practical experience with several of:

  • Agents and agentic workflows

  • Tool/function calling

  • RAG and context engineering

  • Structured outputs

  • MCP or equivalent

  • Evaluation/evals

  • Model selection and routing

• AI is a fundamental part of how you design and build software, not an add-on skill

3. Full-Stack Product Engineering (Core)

• Ability to own a capability end-to-end: user experience → AI/agent layer → services/APIs → data → production

• Hands-on with combinations such as:

  • React / TypeScript

  • Python and/or Java

  • APIs / microservices

  • Databases and cloud

• You needn’t be equally strong in every layer, but you don’t see yourself as only a backend engineer

4. Engineering Fundamentals

• Strong CS and software engineering foundations:

  • Distributed systems, API design, data modeling

  • Concurrency, scalability, reliability

  • Security, debugging and production engineering

• Strong Python engineers, and engineers working across Python/Java/TypeScript, are equally relevant. Deep Core Java, Spring Boot or JVM/GC expertise is not required

5. AI-Augmented Development

• Power user of Claude Code, Cursor, Codex, GitHub Copilot or equivalent

• Can demonstrate how AI has materially changed your productivity and the scope of what you personally build

• Can understand, review, challenge and debug what AI generates; prompting skill alone is not a substitute for engineering capability

6. Production AI (Strongly Preferred)

• Has shipped AI capabilities to production, beyond prototypes, chatbots or POCs

• Can speak concretely about:

  • Evaluation, hallucination and reliability

  • Observability and latency

  • Token/model cost

  • Security, guardrails and scale

7. Education

• Bachelor’s or Master’s degree in Computer Science / Engineering or a related field, or equivalent practical experience

Good to Have

• Founding or early engineer at an AI startup, or experience on an AI platform team

• Built eval harnesses, golden datasets or regression tests for LLM behaviour

• Public evidence of building: GitHub, side projects, write-ups or talks in the AI space

• Exposure to enterprise SaaS, procurement, finance or workflow domains

What This Role Is NOT

• Not an “AI-assisted” backend role where AI only helps you write conventional code faster

• Not a Data Scientist or pure research role

• Not a prompt-engineering-only or POC/chatbot role

• Not a people management-only role

This is a hands-on engineering role for builders who can take a new S2P capability from user problem to production, using AI both in how they build and in what they ship.