Course Objective & Scope
In enterprise workflows, real value comes from turning unstructured contracts, invoices, support tickets, regulatory notices and technical reports into validated database fields and JSON records. This programme covers multi-class, multi-label and hierarchical document classification, supervised and zero-shot Named Entity Recognition (DeBERTa NER, GLiNER), relation and event extraction, hybrid rule-model pipelines (spaCy EntityRuler) and schema-enforced structured extraction with Pydantic.
Layout-aware PDF and table parsing (Docling), entity normalisation/linking and field-level precision/recall evaluation are practised hands-on on enterprise-style documents.
Who Should Attend?
- Data scientists and ML engineers building Intelligent Document Processing (IDP) and text analytics pipelines
- Software teams automating legal, finance, insurance, procurement and customer-support workflows
- Engineers enriching search indices and knowledge bases with structured metadata
- Prerequisite: Python and basic NLP familiarity
Key Takeaways
- Build hierarchical and multi-label classifiers across large enterprise taxonomies
- Extract custom domain entities (amounts, clauses, part numbers, counterparties) using DeBERTa, GLiNER and rule patterns
- Link extracted entities via Relation Extraction into structured table rows and knowledge triples
- Guarantee schema-valid JSON outputs using Pydantic validators and constrained extraction
- Set field-level confidence thresholds to route ambiguous extractions to human-in-the-loop review
Curriculum
Document Triage and Hierarchical / Multi-Label Classification
- Single-label, multi-label and hierarchical taxonomy architectures
- Zero-shot NLI/SetFit cold start and per-class threshold calibration
Enterprise Entity Extraction: Transformer NER and GLiNER
- Token vs span-based NER and handling nested domain entities
- Zero- and few-shot custom entity extraction with GLiNER
Hybrid Rule + Model Pipelines and Entity Normalisation
- Combining spaCy EntityRuler, regex and gazetteers for structured codes and dates
- Canonicalising and linking extracted mentions to master database IDs
Relation Extraction and Schema-Validated JSON Outputs
- Linking entity pairs (party-obligation, item-price, asset-fault)
- Enforcing strict Pydantic schemas and business validation rules on outputs
Layout-Aware PDF Parsing and End-to-End Evaluation
- Preserving reading order and tables in complex PDFs with Docling
- Exact vs partial span metrics and straight-through processing (STP) rate tuning
Other Courses in This Category
-
Enterprise search beyond keyword matching: Sentence-Transformers (bi-encoders), multilingual embeddings, BM25/SPLADE + dense hybrid retrieval, Reciprocal Rank Fusion, Cross-Encoder reranking, Qdrant/pgvector and NDCG@k evaluation.
- Live Course
- 5 modules
-
Deep Learning with PyTorch
Comprehensive deep learning training with PyTorch 2.x covering tensor operations, autograd, CNNs, Transformers, LLM fine-tuning (LoRA/QLoRA), diffusion models, GNNs, and production deployment.
- Live Course
- 12 modules
-
Neural networks, forward pass, backpropagation, activation and loss functions, AdamW, dropout and batch normalisation, giving engineers the skills to build and train their first deep learning models in PyTorch.
- Live Course
- 5 modules
-
Stable, large-scale deep learning covering BF16/FP16 mixed precision, multi-GPU training with DDP/FSDP, self-supervised and contrastive pretraining, metric learning, knowledge distillation and GPU profiling.
- Live Course
- 5 modules
-
From ResNet and ConvNeXt to Vision Transformers (ViT), Swin, hybrid backbones, State Space Models (Mamba), Mixture-of-Experts (MoE) and multimodal fusion: selecting and designing architectures for your problem and hardware.
- Live Course
- 5 modules
-
Self-attention mechanics, Multi-Head and Grouped-Query Attention, RoPE positional encodings, encoder/decoder/encoder-decoder families, KV caching and FlashAttention, implemented from scratch in PyTorch and with Hugging Face.
- Live Course
- 5 modules
Get in touch
