Senior LLM Engineer: Production AI, RAG & Agents (Remote)

BearPlex, Inc; Lahore; Remote; PKR 1,800,000 - 4,200,000; Full time

Benefits offered by this job

Learning budget

Fully remote in Pakistan

Job summary

BearPlex, Inc. is building AI systems that survive enterprise use. You will take LLM-powered features from a working idea to productionDeployment that is observable, evaluated, and safe to deploy for clients who demand proof.

You will engineer LLM applications, design agent loops with tool use and memory, and ship RAG/knowledge pipelines with audit trails so answers are traceable to sources. You will lead evaluation, open-weight model work, and production discipline across the stack.

Qualifications

• Shipped model-powered features in production, keeping systems running for real users.
• Strong Python skills with clean, maintainable code in ML/data tooling.
• Agent and RAG fluency with tool use, retrieval, and context management.
• An evaluation mindset: you measure impact before shipping changes.
• Hands-on with open-weight models (Hugging Face), fine-tuning and serving.
• Production discipline: cost, latency, observability, and auditability matter.

Responsibilities

• Engineer LLM applications and features on frontier models.
• Build autonomous agents with tool use, memory, and guardrails.
• Ship RAG and knowledge systems with audit trails and traceable sources.
• Make evaluation the backbone with offline/online measurement.
• Fine-tune open-weight models and run RLHF when needed.
• Run models in production with latency, caching, and monitoring.
• Support sovereign and on-prem AI for client environments.

Skills

Shipped LLM features

Python

Agent and RAG fluency

Evaluation mindset

Open-weight models

Production discipline

Tools

Hugging Face

PostgreSQL

Docker

Kubernetes

Node.js

TypeScript

Job description

BearPlex, Inc. is building AI systems that survive enterprise use. You will take LLM-powered features from a working idea to productionDeployment that is observable, evaluated, and safe to deploy for clients who demand proof.

You will engineer LLM applications, design agent loops with tool use and memory, and ship RAG/knowledge pipelines with audit trails so answers are traceable to sources. You will lead evaluation, open-weight model work, and production discipline across the stack.