Our client my digital people is hiring a AI Product Engineer in Lahore.

We're looking for a hands-on builder who can go from idea to shipped product using AI — websites and apps, business automations, creative/content generation, and voice agents — and who also has real depth in LLMs and agent tuning under the hood.
What you'll do

Build websites and apps rapidly using AI-assisted development tools (Claude Code, Cursor, or similar), shipping real, working products rather than prototypes
Design and build multi-agent workflows using the Claude Agent SDK, LangChain, LangGraph, CrewAI, or similar — with tool use, state persistence, and self-correction on failure
Build business automations and workflows using n8n (and similar platforms), connecting agents to internal tools, data, and third-party APIs
Build and maintain creative/content generation pipelines — automated image and video generation (e.g. Stable Diffusion, ComfyUI, or similar) for marketing, product, or content use cases
Design, build, and ship production-grade voice agents (Vapi, ElevenLabs, or similar) for real business scenarios — not demos: handling live calls, edge cases, and failure recovery reliably
Fine-tune and tune LLMs/agents (LoRA/QLoRA, prompt/system design, eval-driven iteration) to improve accuracy and reliability for specific use cases
Choose the right approach per problem — build vs. no-code, RAG vs. fine-tuning, one big model vs. several small ones — and justify the trade-off

What you bring

Strong Python; solid grasp of how LLMs/transformers actually behave, not just API usage
Hands-on experience with LangChain, LangGraph, or the Claude Agent SDK, plus agent design patterns (ReAct, Reflection)
Proven ability to ship real websites/apps using AI-assisted coding tools
Practical experience with n8n (or similar) for business automation
Experience building and deploying image/creative generation pipelines
Real production experience with voice agents — ideally something live, handling actual customers or business processes, not just a proof of concept
Practical LLM/agent fine-tuning experience (LoRA/QLoRA, PyTorch or Hugging Face) and judgment on when it beats RAG or prompting
Bachelor's in CS/Engineering/related field, or equivalent practical experience with 2+ years of experience

Nice to have
Vector DBs · MCP · observability tools · real-time voice/telephony integrations · responsible AI/governance familiarity
Work Location: In person

Habilidades requeridas

Python programmingDeep understanding of LLMs/transformers behaviorLLM and agent fine-tuning (LoRA/QLoRA; eval-driven iteration)PyTorch and Hugging Face model toolingLangChain / LangGraph / Claude Agent SDK hands-on usageAgent design patterns (ReAct, Reflection)Designing and building multi-agent workflows (tool use, state persistence, self-correction)Building and shipping websites/apps using AI-assisted coding tools (Claude Code, Cursor)Business automation with n8n or similar and API/integration workImage and video generation pipeline development (Stable Diffusion, ComfyUI, etc.)Designing and deploying production-grade voice agents and real-time voice/telephony integrations (Vapi, ElevenLabs)Architectural decision-making (RAG vs fine-tuning, build vs no-code, model selection trade-offs)Vector databases (nice-to-have)Responsible AI / governance familiarity (nice-to-have)Observability and monitoring tools (nice-to-have)

Empleos similares