AIML Engineer - Senior/Lead

LOWCODEMINDS is a #PeopleFirst company that adapts wholistic well-being as the only true measure of its growth. A workplace full of happy people who deeply care about customers and deliver business outcomes that matter.

Lead/Senior : AI/ML Engineer

Responsibilities:

Train and deploy large language models (LLMs) using techniques such as vector databases (e.g., Faiss, pgvector, Pinecone, Milvus), embeddings (e.g., Gensim, spaCy), transformers (e.g., Hugging Face Transformers, AllenNLP), and natural language processing (NLP) algorithms and techniques.

Optimize LLM performance by tuning hyperparameters and using techniques such as knowledge distillation and fine-tuning

Evaluate and improve accuracy of semantic and hybrid search

Develop and test data pipelines for training and evaluating LLMs using machine learning frameworks such as TensorFlow, PyTorch, and JAX.

Develop and test data extraction from various data sources, pre-process, transform and loading into vector databases via different chunking strategies.

Collaborate with the software development team to integrate LLMs into the chatbot using a chatbot development framework (e.g., Rasa, Rasa NLU, Botpress).

Monitor and maintain LLM performance in production using a tool for monitoring and maintaining LLMs in production (e.g., Weights & Biases, CometML).

Required Tools & Technologies:

Programming languages: Python (Expert).

Gen AI : Expert

Langchain : Expert

DeepEval for evaluating Gen AI applications

GenAI app monitoring using Arize Pheonix

Azure OpenAI : Expert (LLM and Embedding Model)

Using Azure UI to configure Azure services

Azure AI search vectordb

Docker/Dockerfile

Deploying Gen AI solutions to Azure

AWS Textract

Using AWS console to configure services

FASTAPI Python web container for deploying and exposing RESTful services

Qualifications

Degree in computer science, machine learning, or a related field

5+ years of relevant experience (10 years overall experience in the IT industry).

Experience with training and deploying LLMs using the above tools and techniques.