AI ML Engineer | Offshore

Job Summary:

Join our dynamic and innovative team at Alter Domus as an AI & ML Engineer, where you will play a pivotal role in designing, building, and maintaining our AI platform and ML pipelines. Working closely with data analysts, data scientists, and various business stakeholders, you’ll ensure data is reliable, efficient, and accessible in a scalable manner.

What You’ll Do:
 

Build & Prototype

• Assemble off-the-shelf LLMs (Azure OpenAI, AWS Bedrock, Anthropic Claude, Meta LLaMA APIs) into Alter Domus’ products and agentifying our workflows
  • Rapidly prototype in low-code/visual studios (Firebase Studio, v0, Lovable, etc) and with agentic frameworks (LangChain, AutoGen, MCP/A2A).
• Automate & Integrate
  • Connect your assistants to real systems via MCP, APIs, webhooks and microservices.
• Code with AI Helpers
  • Leverage GitHub Copilot, Cursor, Windsurf, Cline and Q Developer to accelerate script, API and workflow development.
  • Write unit/integration tests and help maintain our CI/CD pipelines.
• Live Projects You’ll Continue to Develop
  • domusAI:  An adaptive AI assistant platform for AD employees that learn user preferences through a dynamic memory management system. It provides personalized support via agentic LLM chats and integrated tool capabilities. The system features semantic document search, contextual understanding, MCP integration, custom agent creator, and secure information retrieval.
  • domiAI: Our flagship virtual assistant powered by an integrated data and AI platform that transforms client-specific information into vectorized knowledge bases and structured data repositories, enabling real-time generation of meaningful insights, comprehensive reports, search, analytics, and predictions through an intuitive conversational interface.
  • domusDocs: A document extraction engine to extract intelligence and process business documents leveraging advanced GenAI techniques
  • Working on Alter Domus Knowledge Based Agents and vectorizing our corpus of data
  • Multi-Agentic Framework: The architecture that lets specialized AI agents coordinate on complex workflows, build MCPs—your chance to see multi-agent collaboration in action.

Learn & Collaborate

• Learn: Work with an agile team of data scientists, engineers, and AI specialists to master cutting-edge techniques in data analysis, prompt engineering, and AI implementation.
  • Collaborate: Rapidly prototype solutions, iterate based on client feedback, and deploy customized solutions that address specific business challenges while continuously expanding your technical capabilities and domain expertise.

Support & Iterate

• Help define test plans, troubleshoot issues and tune prompts for accuracy and performance.
  • Gather feedback from business stakeholders and iterate quickly on your prototypes.
     

Who You Are:

• Tech Enthusiast: You’ve built at least three AI end-to-end projects deployed in production
• Curious Learner: You’re eager to continue exploring new LLM services, low-code studios and agentic frameworks.
• Collaborative Communicator: You explain technical ideas clearly in English and enjoy working closely with SMEs and teammates.
• Adaptive: You are quickly able to adapt to the development and deployment Golden Path internal to Alter Domus
• Self-Starter: You solve problems proactively using online resources, community forums—or even autonomous agents of your own creation.

Technical Requirements:

•  5+ years of total experience with relevant experience of 3+ years in Python and related data science libraries (NumPy, Pandas, PyTorch, TensorFlow)
•  Experience creating agentic AI workflows using Lang graph, Autogen, CrewAI or other agent libraries
•  Experience with vector databases (Pinecone, Qdrant, or similar)
•  Proficiency with REST APIs and microservice architecture
•  Familiarity with AWS cloud platform
•  Experience with containerization (Docker) and orchestration tools (Kubernetes)
•  Strong understanding of NLP concepts and LLM implementation
•  Experience with prompt engineering and fine-tuning language models