About ARAs Client
ARAs Client is a global leader in strategy, technology consulting and digital services, enabling clients in more than 100+ countries to implement cutting edge solutions in their transformation journey. They helps businesses become digital-first and digital-ready. The company has a global presence with 600000+ employees ( 200000+ in India).
Role Summary
We are seeking a Lead AI / Machine Learning Architect to own the technical design and delivery of advanced AI systems, spanning Agentic AI, Generative AI (LLMs), and classical Machine Learning / Computer Vision.
Key Responsibilities
- Define end-to-end architecture for LLM and agent-based systems, including orchestration, tool use, safety guardrails, and observability.
- Architect and optimize search & retrieval systems (BM25 + vector embeddings) and RAG pipelines, including relevance tuning and evaluation.
- Oversee web scraping and data extraction pipelines using Playwright/Selenium/Trafilatura with structured normalization (JSON, Parquet, schema validation).
- Establish evaluation strategies (offline/online), defining quality, latency, and cost KPIs; integrate LLM testing frameworks (e.g., DeepEval).
- Guide data governance, privacy, and PII handling, working closely with MLOps and platform teams.
Must-Have Qualifications
- 7.5+ years of hands-on experience in Machine Learning / AI engineering
- Proven experience architecting and deploying production-grade ML and LLM systems
Strong expertise in:
- PyTorch and Hugging Face ecosystem (Transformers, Datasets, Sentence-Transformers/SBERT)
- LLMs (tokenization, context windows, inference optimization, fine-tuning/PEFT, RAG)
- Search & retrieval: Elastic/OpenSearch, FAISS / Pinecone / pgvector
- ML frameworks: scikit-learn, spaCy, LightGBM, Optuna, SHAP
- Strong understanding of evaluation strategies, explainability, and responsible AI
- Excellent written and verbal communication skills
- 15 years of full-time education (mandatory)
