Sparktek
Job Title: ML Engineers - with LLM GenAI (3 Resources)
Responsibilities
Write efficient machine learning workflows and pipelines
Training pipeline - Ingest/Preprocess/Vectorize and index data Inference pipeline - AI Guided workflow to respond to user requests or provide proactive recommendations
Measure Model metrics to evaluate the performance of the model (Predictive & Gen AI metrics) - ( F1 Score, Faithfulness, Answer Relevancy etc. ) Evaluate and compare model performance for specific tasks to enable picking the optimal models Apply Guardrails for both inputs and outputs Identify areas of improvements for machine learning pipelines and models and implement them Collaborate with Data scientists, MLOps/Devsecops and UX engineers to implement the solution Collaborate with cross functional teams to translate business requirements into ML solutions Document and communicate, ml workflows, algorithms and solutions to technical and non-technical stakeholders AI application experience with Network Domain solutions would be a plus Skillset
ML Engineer with 4-6 Years of relevant experience Excellent experience with Python NLP libraries (NLTK, gensim, spacy etc.) Good experience with deploying production models for Classification / Regression tasks Working experience with LLM Application frameworks like langchain Working experience with LLM Data and pre-processing frameworks like Llamaindex and unstructured.io Unstructured | The Unstructured Data ETL for Your LLM Unstructured helps you get your data ready for AI by transforming it into a format that large language models can understand. Easily connect your data to LLMs. unstructured.io Experience with applying guardrails for machine learning workflows. Experience with nemo-guardrails for LLM and microsoft presidio or any other PII removal frameworks Working experience with open source and commercial LLM models Working experience in benchmarking embedding models (both open and closed source) for vectorization and indexing Working experience on integrating with any Vector Databases (elastic, quadrant etc.) Experience with any MLOps frameworks (open source or commercial) would be a plus Experience with any cloud based AI platforms (Sagemaker, Vertex AI etc.) Experience with HungginFace libraries (SentenceTransformers etc.) Adept at prompt engineering - Using the right prompts that maximises the accuracy of the Models response Excellent experience with Python and the related ecosystem for package management Rust, Go or C for compute efficient workflows like efficient model serving would be a plus Excellent knowledge of Notebook environment (Jupyter, VS Code IDE) for experimentation and development Good experience with Git platforms and development workflows (Github, Gitlab etc.)
Responsibilities
Write efficient machine learning workflows and pipelines
Training pipeline - Ingest/Preprocess/Vectorize and index data Inference pipeline - AI Guided workflow to respond to user requests or provide proactive recommendations
Measure Model metrics to evaluate the performance of the model (Predictive & Gen AI metrics) - ( F1 Score, Faithfulness, Answer Relevancy etc. ) Evaluate and compare model performance for specific tasks to enable picking the optimal models Apply Guardrails for both inputs and outputs Identify areas of improvements for machine learning pipelines and models and implement them Collaborate with Data scientists, MLOps/Devsecops and UX engineers to implement the solution Collaborate with cross functional teams to translate business requirements into ML solutions Document and communicate, ml workflows, algorithms and solutions to technical and non-technical stakeholders AI application experience with Network Domain solutions would be a plus Skillset
ML Engineer with 4-6 Years of relevant experience Excellent experience with Python NLP libraries (NLTK, gensim, spacy etc.) Good experience with deploying production models for Classification / Regression tasks Working experience with LLM Application frameworks like langchain Working experience with LLM Data and pre-processing frameworks like Llamaindex and unstructured.io Unstructured | The Unstructured Data ETL for Your LLM Unstructured helps you get your data ready for AI by transforming it into a format that large language models can understand. Easily connect your data to LLMs. unstructured.io Experience with applying guardrails for machine learning workflows. Experience with nemo-guardrails for LLM and microsoft presidio or any other PII removal frameworks Working experience with open source and commercial LLM models Working experience in benchmarking embedding models (both open and closed source) for vectorization and indexing Working experience on integrating with any Vector Databases (elastic, quadrant etc.) Experience with any MLOps frameworks (open source or commercial) would be a plus Experience with any cloud based AI platforms (Sagemaker, Vertex AI etc.) Experience with HungginFace libraries (SentenceTransformers etc.) Adept at prompt engineering - Using the right prompts that maximises the accuracy of the Models response Excellent experience with Python and the related ecosystem for package management Rust, Go or C for compute efficient workflows like efficient model serving would be a plus Excellent knowledge of Notebook environment (Jupyter, VS Code IDE) for experimentation and development Good experience with Git platforms and development workflows (Github, Gitlab etc.)