

Twine
AI Engineer – Freelancer
⭐ - Featured Role | Apply direct with Data Freelance Hub
This role is for an AI Engineer – Freelancer, focusing on optimizing an AI data pipeline in a serverless AWS environment. Contract length is unspecified, with a pay rate of "unknown." Key skills include Python, AWS, multi-modal embeddings, and experience with Pinecone and OpenSearch.
🌎 - Country
United States
💱 - Currency
$ USD
-
💰 - Day rate
Unknown
-
🗓️ - Date
October 23, 2025
🕒 - Duration
Unknown
-
🏝️ - Location
Unknown
-
📄 - Contract
Unknown
-
🔒 - Security
Unknown
-
📍 - Location detailed
United States
-
🧠 - Skills detailed
#AI (Artificial Intelligence) #ML (Machine Learning) #AWS (Amazon Web Services) #Langchain #Databases #OpenSearch #Batch #Data Pipeline #Python #Scala
Role description
Join a project focused on advancing an AI-powered influencer marketing platform by optimizing and scaling its data pipeline. The primary objective is to enhance the performance, cost-efficiency, and recall quality of image-to-text and embeddings workflows. You will be responsible for refining batching, caching, concurrency, and queuing within a serverless AWS environment, as well as tuning models such as InternVL and Qwen using RunPod and Replicate. The role also involves improving semantic recall and retrieval quality for influencer content leveraging Pinecone and OpenSearch, and exploring multi-modal embeddings to boost search accuracy. There is potential to expand the system into Retrieval-Augmented Generation (RAG) and agentic AI workflows for automated insights and recommendations.
Responsibilities
• Take ownership of the existing AI data pipeline, focusing on performance and cost optimization
• Optimize image-to-text and embeddings pipelines for speed, efficiency, and recall quality
• Implement and improve batching, caching, concurrency, and queuing in a serverless AWS environment
• Tune and deploy models (e.g., InternVL, Qwen) via RunPod and Replicate
• Enhance semantic recall and retrieval quality using Pinecone and OpenSearch
• Experiment with multi-modal embeddings for improved search accuracy
• Contribute to the development of RAG and agentic AI workflows for automated insights and recommendations
Skills And Requirements
• Proven experience optimizing AI/ML pipelines, particularly image and embedding workflows
• Strong proficiency in Python and AWS serverless architecture
• Hands-on expertise with multi-modal embeddings and semantic search
• Familiarity with vector databases (e.g., Pinecone, OpenSearch)
• Experience with LangChain, RAG, and related frameworks
• Ability to work independently and deliver robust, scalable solutions
• Excellent problem-solving and communication skills
About Twine
Twine is a leading freelance marketplace connecting top freelancers, consultants, and contractors with companies needing creative and tech expertise. Trusted by Fortune 500 companies and innovative startups alike, Twine enables companies to scale their teams globally.
Our Mission
Twine's mission is to empower creators and businesses to thrive in an AI-driven, freelance-first world.
Join a project focused on advancing an AI-powered influencer marketing platform by optimizing and scaling its data pipeline. The primary objective is to enhance the performance, cost-efficiency, and recall quality of image-to-text and embeddings workflows. You will be responsible for refining batching, caching, concurrency, and queuing within a serverless AWS environment, as well as tuning models such as InternVL and Qwen using RunPod and Replicate. The role also involves improving semantic recall and retrieval quality for influencer content leveraging Pinecone and OpenSearch, and exploring multi-modal embeddings to boost search accuracy. There is potential to expand the system into Retrieval-Augmented Generation (RAG) and agentic AI workflows for automated insights and recommendations.
Responsibilities
• Take ownership of the existing AI data pipeline, focusing on performance and cost optimization
• Optimize image-to-text and embeddings pipelines for speed, efficiency, and recall quality
• Implement and improve batching, caching, concurrency, and queuing in a serverless AWS environment
• Tune and deploy models (e.g., InternVL, Qwen) via RunPod and Replicate
• Enhance semantic recall and retrieval quality using Pinecone and OpenSearch
• Experiment with multi-modal embeddings for improved search accuracy
• Contribute to the development of RAG and agentic AI workflows for automated insights and recommendations
Skills And Requirements
• Proven experience optimizing AI/ML pipelines, particularly image and embedding workflows
• Strong proficiency in Python and AWS serverless architecture
• Hands-on expertise with multi-modal embeddings and semantic search
• Familiarity with vector databases (e.g., Pinecone, OpenSearch)
• Experience with LangChain, RAG, and related frameworks
• Ability to work independently and deliver robust, scalable solutions
• Excellent problem-solving and communication skills
About Twine
Twine is a leading freelance marketplace connecting top freelancers, consultants, and contractors with companies needing creative and tech expertise. Trusted by Fortune 500 companies and innovative startups alike, Twine enables companies to scale their teams globally.
Our Mission
Twine's mission is to empower creators and businesses to thrive in an AI-driven, freelance-first world.






