Sr. NLP (Natural language processing) Engineer (01 Position)
Lahore, Pakistan
Job Description :
Research on creating state-of-the-art, scalable, and self-learning systems for our Business KPIs and their sub-components.
Design, train and tune a variety of Deep Neural Networks / Machine Learning Models.
Create models from varying levels of Data Quality and Quantity (Imbalanced, Multi-Domain and/or Multi-Class, Unlabeled) - here your ability to creatively apply Domain Knowledge will be important, such as driving existing model implementations from Supervised to Semi-Supervised Learning.
Help define and drive improvements related to Architectural Designs, Dialogue-Schemas, Data Acquisition, Feature Transformations and Model Evaluations.
Standardize and Automate Annotation Practices for efficient utilization for the acquired data.
Formulate general purpose solutions configurable to accommodate custom requirements for different clients.
Formulate general purpose solutions configurable to accommodate custom requirements for different clients.
Perform Data and Error Analysis in order to Improve Models and Understand their Shortcomings.
Lead all NLP-based projects for AI involving NLP Engineers and Data Analysts.
Communicate and Work effectively with Team Members and Other Teams (on-site and/or remotely).
We are looking for:
BS(CS) or MS(Computational Linguistics)
5 years’ or more
Skills:
Expert knowledge with a scripting language (preference: Python), an object-oriented language (e.g. C++, Java) and a query language (preference: SQL).
Expert knowledge and Experience of building ML Pipeline Architectures and the general lifecycle of an NLP Project.
Experience and/or demonstrable knowledge of manual and/or automated data analysis and its techniques for extracting meaningful insights and be able to communicate them effectively.
Expert knowledge and Experience of working on conventional ML/DL problems like: Supervised Classification and Unsupervised Clustering, as well as some optional knowledge about RL (Reinforcement Learning).
Expert knowledge and Experience with classic NLP Feature-Extraction techniques such as: n-grams, part-of-speech tagging, semantic distance metrics, search indexing, and corpus analysis etc.
Expert knowledge and Experience of working on conventional Computational Linguistics problems such as: Syntactic Processing, Word Sense Disambiguation, Context-Free Grammars, Lexical Semantics, Quantification and Plurality, etc.
Expert knowledge and Experience of working in conventional NLP problems like: Informational Retrieval, Relevance Ranking and Search, Question-Answer Generation, Natural Language Inference, Topic Modeling etc.
Experience and/or demonstrable knowledge of working in Conversational AI: particularly in its classical NLP problems like: Automatic Speech Recognition (ASR), Machine Translation, Chatbot Response Generation, Frame-wise Spoken Dialogue-Intent Recognition, Dialogue-Act Classification and Slot-Labeling, Turn Allocation and Dialogue State Tracking, etc.
Optional experience working with both Audio and Text Datasets on spoken language problems.
Experience working with Deep learning and ML API’s such as Tensorflow, PyTorch, XGboost, CatBoost, etc as well as a firm grip over modern state-of-the-art and otherwise NLP architectures such as BERT, GPT3, Reformer, RNNs, CNNs, LSTMs etc and vector space models such as GloVe, Word2vec, USE etc.
Ability to communicate technical concepts and solutions at a level appropriate for technical/non-technical audiences.
Have a decisive personality and ability to utilize past experience to derive ETAs, timelines, resource consumption and realistic result expectations.
Ability to define and structure project lifecycles: conceptualization, architectural and database design, skeletal construction, task delegation to inter and intra team developers, milestone deliverables and possible improvement tracks.
Ability to explain and justify research and analysis done for a cause and deliver acquired results to the higher management.
Ability to point out fallacies / pitfalls in implemented model architectures, algorithms and data processing.
Ability to understand the goals of the project and the end user and set the right expectations with higher management after analyzing AI and business constraints.
Experience working with Numerical Computing Tools like: NumPy, SciPy – using the standard Pandas API.
Experience working with ML/DL frameworks like: Scikit-Learn, Tensorflow, Keras, PyTorch, etc.
Familiarity with utilizing hardware acceleration concepts and tools such as Numba, PyCuda, TFX Runtime, etc.
Familiarity with Model Serving APIs like: Tensorflow Serving, TorchServe, Django & Flask, etc.
Good to haves: Familiarity with Model Serving APIs like: Tensorflow Serving, TorchServe, Django & Flask and Familiarity with Utilizing hardware acceleration concepts and tools such as Numba, PyCuda, TFX Runtime, etc.