Ai Architect

Summary

Join Expedite Commerce as an AI Architect and design, build, and fine-tune NLP models and large language model (LLM) agents using AWS Sagemaker and Bedrock. You will play a key role in creating intuitive and efficient model designs enhancing user experiences and business processes. This position requires strong design skills, hands-on coding expertise in Python, advanced proficiency in LLM agent design and development, and exceptional debugging capabilities. You will oversee all aspects of model architecture, data pipeline integration, and metrics interpretation. The role demands a deep understanding of LLM technologies and a commitment to continuous learning. This is a fully remote opportunity with benefits including health insurance, PTO, paid professional training, and a strong onboarding program.

Requirements

• Possess strong design skills
• Possess hands-on coding expertise
• Possess advanced proficiency in Python development
• Possess specialized knowledge in LLM agent design and development
• Possess exceptional debugging capabilities
• Possess strong skills in deploying models and agents on cloud platforms, particularly AWS, and implementing serverless architectures
• Possess expertise in debugging and fixing issues related to LLMs
• Possess excellent written and verbal communication skills in English

Responsibilities

• Oversee all aspects of model architecture, data pipeline integration, and metrics interpretation
• Design scalable and optimized solutions for training, retraining, deploying, scheduling, monitoring, and improving NLP models and LLM agents
• Conceptualize and design robust NLP solutions and LLM agents tailored to specific business needs, with a focus on user experience, interactivity, latency, failover and functionality
• Write, test, and maintain clean, efficient, and scalable code for NLP models and LLM agents, with a strong emphasis on Python programming
• Develop and fine-tune LLM agents, leveraging advanced techniques in Deep Learning and Transformer architectures, including models like BERT, GPT, Whisper, ChatGPT, and other generative models
• Monitor, optimize LLM agents, implementing model explainability, handling model drift, and ensuring robustness
• Read, comprehend, and implement LLM research papers into practical solutions
• Stay abreast of the latest academic and industry research to apply cutting-edge methodologies and techniques
• Proactively identify, diagnose, and resolve issues related to LLM models, including model inaccuracies, performance bottlenecks, and system integration problems
• Utilize debugging tools and techniques to troubleshoot complex problems in model behavior, data inconsistencies, and deployment errors
• Stay updated with the latest advancements in LLM technologies, experimenting with new techniques and tools to enhance agent capabilities and performance
• Adapt to unlearn outdated practices, patterns, technologies and quickly learn and implement new technologies & papers as the ML world evolves
• Maintain a proactive approach to staying current with emerging trends and technologies in Agent based solution (Text & Multi Modal)
• Demonstrate the ability to design complex systems including LLM agents, with experience in architecting solutions from conceptualization to deployment
• Demonstrate extensive experience in coding of LLM agents, with advanced proficiency in Python (3.10+), including knowledge of frameworks such as PyTorch, or similar
• Demonstrate 4-6 years of experience in fine-tuning LLMs and deploying LLM agents, including practical experience with AWS Bedrock, OpenAI Function Calling, Anthropic Function Calling, CrewiAI, Meta GPT framework, and other relevant platforms
• Demonstrate a proven track record of developing high-quality, efficient Python code, including experience with advanced Python features and best practices
• Demonstrate experience with integrating open-source and commercial LLM models and LLM agents, including developing and evaluating prompt engineering techniques
• Demonstrate a deep understanding of multi-modal model architecture, experience with AWS Bedrock agent models, and practical experience in fine-tuning models for specific use cases
• Demonstrate strong skills in deploying models and agents on cloud platforms, particularly AWS, and implementing serverless architectures (Utilizing AWS Lambda, Kinesis, SQS, DDB, Bedrock, OpenAI API, S3, Step Function)
• Demonstrate expertise in debugging and fixing issues related to LLMs, including identifying root causes of errors, resolving discrepancies in model outputs, and optimizing system performance
• Demonstrate excellent written and verbal communication skills in English, with the ability to present technical concepts clearly with team and clients
• Demonstrate experience in CI/CD pipeline using AWS CodePipeline, CodeBuild, and CodeDeploy for automated testing and deployment of NLP Solution
• Demonstrate a demonstrable portfolio of LLM projects and LLM agent developments, with contributions to public forums like Kaggle, open-source projects, and publications in technical forums
• Demonstrate experience in understanding latest technologies/papers and implementing the same into the solution
• Demonstrate hands-on experience in CI/CD solution utilizing AWS services (Code Commit, Code Build & Code Pipeline)

Preferred Qualifications

Possess a good understanding of multimodal solution finetune and deployment

Benefits

• Health Insurance
• PTO
• Leave time
• Ongoing paid professional training and certifications
• Fully Remote work Opportunity
• Strong Onboarding & Training program