Prompt Engineer — AI & CX Solutions

Pronix Inc; Hyderabad; On-site; INR 1,800,000 - 2,800,000; Full time

Job summary

Pronix Inc in Hyderabad seeks an experienced Prompt Engineer to design, test and govern production prompt systems for enterprise CX agents and backend AI integrations.

You will translate industry use cases into reusable architectures, ensure safety and escalation, and drive high-quality task completion across voice, chat and email channels while collaborating with architects, engineers and client teams through production monitoring.

Qualifications

• 4+ years of experience in prompt engineering, conversational AI or applied NLP with production work on enterprise AI agents.
• Ability to write and debug prompts for multistep agent workflows, RAG, structured outputs and function/tool calling.
• Hands-on with leading LLM providers and agent frameworks; design portable prompt patterns across vendors.
• Strong knowledge of prompt evaluation, safety controls, hallucination reduction, PII handling, versioning and observability.

Responsibilities

• Design, test, version and optimize system prompts and conversation patterns for CX agents across voice, chat, messaging, email and agent-assist channels.
• Translate industry use cases and operating policies into reusable prompt architectures including context management, escalation and human handoff.
• Build prompt systems for backend AI agents handling documents, workflow automation and tool integrations.
• Create evaluation datasets and scorecards to measure quality, safety, latency and cost; use results to improve prompts and routing.
• Collaborate with solution architects, engineers and client teams from discovery to production monitoring and optimization.

Skills

Prompt engineering

Conversational AI

LLM platforms

Prompt testing

Tool calling

Client communication

Tools

Agent frameworks

Enterprise AI platforms

Job description

Design, evaluate and govern production prompt systems for industry-specific CX agents, omnichannel conversations and backend generative AI agents that automate enterprise workflows and integrate with LLMs.

Responsibilities

• Design, test, version and optimize system prompts, tool instructions and conversation patterns for customer-facing AI agents across voice, chat, messaging, email and agent-assist channels.
• Translate industry use cases, customer journeys and operating policies into reusable prompt architectures for CX agents, including context management, escalation, safety and human handoff.
• Build prompt systems for backend generative AI agents that handle documents, workflow automation, structured outputs, enterprise APIs, tool calling and LLM integrations.
• Create evaluation datasets and scorecards for answer quality, groundedness, task completion, safety, latency and cost; use measured results to improve prompts, retrieval and routing.
• Partner with solution architects, conversation designers, engineers and client teams from discovery and prototyping through production monitoring and continuous optimization.

Requirements

• 4+ years in prompt engineering, conversational AI, generative AI, conversation design or applied NLP, including production work on enterprise AI agents.
• Rock-solid ability to write and debug prompts for industry-specific CX use cases, multistep agent workflows, RAG, structured output and function/tool calling.
• Hands-on experience with leading LLM providers, agent frameworks and enterprise AI platforms, with the ability to design portable prompt patterns rather than depend on one vendor.
• Strong knowledge of prompt evaluation, safety controls, hallucination reduction, PII handling, versioning, observability and responsible human escalation.
• Clear client-facing communication and the ability to turn business processes, policies and customer journeys into testable AI behavior.

Expected outcomes

• Ship prompt systems that meet agreed quality, safety and containment targets in production.
• Build reusable prompt libraries and evaluation scorecards that accelerate future CX agent builds.
• Measurably reduce hallucination and escalation failures across the agents you own.