AI Architect/Lead Consultant (NLP / Transformers)

AI Architect/Lead Consultant (NLP / Transformers)

Location: Calgary (Onsite)

Duration: 12 month

Experience: 8–10 Years

Role Overview: We are looking for an experienced AI Engineer (NLP / Transformers) to design and build intelligent systems capable of reading, interpreting, and contextualizing contract documents alongside user inputs

This role focuses on leveraging transformer-based models, embeddings, and retrieval architectures to enable deep document understanding and context-aware reasoning, particularly within domain-specific datasets such as railroad tariff documents

The ideal candidate will bring strong expertise in LLMs, semantic search, and contextual memory frameworks, enabling the system to learn from historical data and apply that knowledge to new documents

Key Responsibilities

• Design and implement transformer-based NLP models for document understanding and interpretation

• Build systems that can ingest, process, and extract insights from contract documents

• Develop embedding-based semantic search and retrieval mechanism

• Implement contextual memory architectures (e.g., RAG, vector databases) to enable learning from previously processed documents

• Enable context-aware reasoning by combining user inputs with historical document knowledge

• Handle and model domain-specific terminology, especially railroad tariff language (e.g., T3 slips)

• Fine-tune and optimize LLMs for domain-specific use cases

• Collaborate with data engineering teams to integrate models into scalable pipelines

• Ensure model accuracy, explainability, and performance optimization

• Conduct experimentation, prototyping, and PoC development

Key Deliverables

• NLP-based document understanding solution (contracts + user inputs)

• Embedding models and semantic search implementation

• Contextual memory / retrieval architecture design

• Fine-tuned transformer models for domain-specific use

• Evaluation metrics, accuracy benchmarks, and validation reports

• PoC artifacts and technical documentation

Required Skills & Qualifications:

• Strong experience with transformers / LLMs (e.g., BERT, GPT-based models)

• Expertise in embeddings and semantic search frameworks

• Proficiency in Python, with hands-on experience in:

--- PyTorch

---TensorFlow

• Experience in NLP for document understanding (contracts, legal text, structured extraction)

• Knowledge of retrieval-based architectures (e.g., RAG, vector DBs like FAISS, Pinecone)

• Experience building contextual memory systems

• Strong understanding of model fine-tuning, evaluation, and optimization

Preferred Qualifications

• Experience working with legal, contract, or tariff-based datasets

• Exposure to railroad/logistics domain data

• Familiarity with LLM orchestration frameworks (e.g., LangChain, LlamaIndex)

• Understanding of data pipelines and production ML systems