Title:  Architect- AI / ML

Area(s) of responsibility

About the Role

We are seeking a hands-on, highly experienced AI Architect with deep expertise in LLMs, Retrieval-Augmented Generation (RAG), fine-tuning, and applied machine learning. You will help build next-generation investor relations and public relations tools powered by AI, including intelligent assistants, content generation, semantic search, and predictive analytics.

Machine Learning + Deep Learning + Generative AI

Key Responsibilities

• Design and implement RAG pipelines that combine proprietary data (e.g., press releases, earnings transcripts) with LLMs to generate accurate, secure outputs.
• Fine-tune open-source LLMs (e.g., LLaMA 2/3, Mistral, Mixtral) using instruction or LoRA-based methods for specific use cases such as branded content generation or question-answering.
• Build and optimize embeddings, Vector search algorithms proficenecy e.g. HNSW, FAISS, IVFFlat and Microsoft -> Semantic Kernel and semantic retrieval systems using tools like sentence-transformers, FAISS, Weaviate, or Pinecone.
• Collaborate with product and content teams to implement intelligent Q&A, press release graders, interview prep tools, and performance analysis assistants.
• Develop traditional ML models for anomaly detection, time-series forecasting, and engagement optimization (Prophet, XGBoost, scikit-learn, Chronos, ARIMA, Microsoft -> Semantic Kernel etc.).
• Integrate multimodal components such as TTS (e.g., ElevenLabs), image recommendation, and social publishing automation.
• Deep Learning in Neural Networks.
• Hands on experience in Tensorflow, Pytorch.
• Ensure secure and scalable deployment of AI systems (on cloud AWS, AWS Bedrock, AWS SegemakerAzure OpenAI, or in-house GPU infrastructure).

Qualifications

• Hands-on experience in ML/AI engineering or applied data science
• Proven experience with LLMs (GPT-4, Claude, LLaMA, Mistral, etc.) and RAG architectures
• Strong coding skills in Python and experience with LangChain, Hugging Face Transformers, OpenAI API, PEFT/LoRA.
• Experience fine-tuning LLMs using open-source frameworks and adapting models with domain-specific data
• Deep understanding of NLP, search, and vector-based retrieval systems
• Experience with time-series modeling, anomaly detection, or ML classification tasks
• Familiarity with DevOps or MLOps tooling for deploying and monitoring LLM-powered applications
• Excellent communication skills and ability to work cross-functionally with product, content, and engineering teams

Preferred

• Experience in financial, PR, or media domains
• Familiarity with embedding alignment, model evaluation, and prompt engineering
• Comfortable working with both cloud APIs (e.g., OpenAI, Azure, Anthropic) and open-source LLMs

Good to have:

Postgressql vector db for embeddings, azure open ai services and c# .net 9 framework for building all business logic and services for our enterprise applications.