Senior Machine Learning Engineer

trcglobal Pakistan
Remote
Apply
AI Summary

Join our team to build the next-generation platform for intelligent document data processing. We're looking for a hands-on ML Engineer to design and develop a hybrid of classic ML and modern GenAI. The ideal candidate will have experience with Python, PyTorch, TensorFlow, and Agentic frameworks like LangChain or LangGraph.

Key Highlights
Design and develop a hybrid of classic ML and modern GenAI
Build orchestration flows using LangGraph or similar frameworks
Tackle complex data problems, including advanced chunking strategies and semantic search
Technical Skills Required
Python PyTorch TensorFlow LangChain LangGraph Pinecone Milvus Weaviate Cloud Run Pub/Sub
Benefits & Perks
Competitive salary
Remote work
European work standards
Global impact
USD stability

Job Description


The Mission 🚀

Let’s be honest: the world is still buried in messy, unstructured, and highly variable document data. We are partnering with a leading Finnish-based client to build the next-generation platform that changes that 

We aren't just processing files. We are building an intelligent system that understands context, reasons through complexity, and automates expertise. We need a hands-on ML Engineer based in Pakistan to join our global efforts. This is a role for someone who lives at the intersection of classic statistical rigor and the cutting edge of Agentic AI.


Location and Compensation 🇵🇰

Role: Fully Remote (Work from anywhere in Pakistan)

Client Location: Helsinki, Finland

Employment Type: On the payroll of a Finland-based Technology Services Provider

Monthly Salary: $1,000 to $3,000 USD (Stable and competitive pay)


What You Will Actually Do 🛠️

You won't be siloed. You will be the bridge between data science and robust engineering:

  • Build the Brain: You will design a hybrid of classic ML, like classification and extraction, alongside modern GenAI. You will fine-tune LLMs and embedding models to handle the nuances of complex multi-page PDFs and tables.
  • Move Beyond the Prompt: We are going deep into Agentic AI. You will build orchestration flows using LangGraph or similar frameworks, creating agents that can use tools, remember context, and evaluate their own performance.
  • Solve the Messy Data Problem: You will tackle the hard stuff, including advanced chunking strategies for mixed formats, semantic search, and building high-performance RAG pipelines.
  • Engineer for Scale: Our stack is GCP native. You will architect microservices and event-driven pipelines using Cloud Run and Pub/Sub that scale effortlessly.


The Skill Stack We Value 🧠

We are looking for a practitioner, not just a researcher. If you have a GitHub full of experiments or a history of shipping models, we want to talk.

The Essentials:

  • The ML Core: Deep proficiency in Python, PyTorch, or TensorFlow, and the math behind the models, such as feature engineering and statistical tuning.
  • The GenAI Edge: Proven experience building RAG systems and working with Agentic frameworks like LangChain or LangGraph. You know how to make LLMs behave in production.
  • The NLP Toolkit: Experience with embeddings, text segmentation, and entity extraction specifically for document intelligence.
  • Data Architecture: Hands-on experience with Vector Databases such as Pinecone, Milvus, or Weaviate, and building scalable microservices.
  • Cloud Fluency: Experience deploying production workloads, ideally within the GCP ecosystem.


Why Join Us? ✨

  • European Work Standards: Experience the Finnish way of working with high trust, clear communication, and a focus on quality over crunch time.
  • Global Impact: Work on world-class AI infrastructure from the comfort of your home in Pakistan while being on the payroll of a European provider.
  • USD Stability: Protect your earnings with a competitive salary paid in USD, offering significant local leverage and financial security.
  • Cutting Edge Stack: We stay at the forefront. If there is a better tool or a more efficient architecture, we want you to find it and implement it.

Subscribe our newsletter

New Things Will Always Update Regularly