Artificial Intelligence Specialist

Work Location: Chennai (Work from Office) Role Overview We’re hiring an NLP/LLM Engineer to design production-ready NLP pipelines, fine-tune open-source LLMs, and integrate LLMs into real-time products. You’ll work with product teams to build classification, NER, summarization, sentiment, and QA systems — from data preprocessing to deployment and monitoring. Responsibilities Design and implement NLP pipelines for classification, NER, summarization, sentiment analysis, and Q&A (RAG). Preprocess and structure large-scale textual datasets using SpaCy, NLTK, or custom tokenizers. Fine-tune and optimize open-source models (BERT, T5, GPT-style) using Hugging Face Transformers. Integrate LLM APIs (OpenAI, Anthropic/Claude, Cohere, Hugging Face) into real-time applications and microservices. Develop prompt strategies and evaluate prompt robustness; implement few/zero-shot approaches where relevant. Build retrieval + generation (RAG) pipelines with vector DBs (FAISS, Weaviate, Pinecone). Deploy model endpoints (Flask/FastAPI/Streamlit), containerize with Docker, and collaborate on cloud deployment. Monitor model performance, data drift, and user feedback; iterate based on metrics and qualitative signals. Collaborate with product, infra, and QA to embed models safely and responsibly. Required skills & experience Strong Python skills and experience with SpaCy, Transformers (Hugging Face), and NLTK. Hands-on experience fine-tuning/training Transformer models (BERT, T5, GPT variants). Experience integrating LLM APIs and working with prompt engineering. Familiarity with embeddings, attention, tokenization, and vector similarity (cosine). Experience with vector DBs such as FAISS, Weaviate, or Pinecone. Comfortable with Docker, Git, and REST API development (Flask/FastAPI). Practical knowledge of evaluation metrics (BLEU, ROUGE, perplexity, cosine similarity). For Sr. Roles: In addition to the above, Track record of shipping multiple production LLM/NLP systems end-to-end. Strong system design skills (latency/cost tradeoffs, multi-tenant inference). Mentoring/leadership experience and ownership of ML lifecycle (MLOps). Preferred / nice-to-have Experience with LangChain, LlamaIndex, PromptLayer, or similar LLMOps tooling. RAG workflows and building retrieval pipelines at scale. Experience in cloud deployment on AWS/GCP/Azure. Contributions to open-source projects or research publications. About Ltd, a Great Place to Work Certified Wed-Tech company listed on the BSE and NSE, is a leader in matchmaking and wedding services. Its flagship brand, Bharat Matrimony, is the largest and most trusted matrimony brand (Brand Trust Report 2014) spanning customers across the entire economic strata. The company serves over 6 million members with the support of 3,000+ associates and a network of 125+ company-owned retail outlets. Key Offerings Matchmaking Services: BharatMatrimony and 200+ matrimony brands segmented on religion, languages, communities, and occupation. Jodii: An exclusive matchmaking app for non-degree holders in 9 Indian languages Assisted Service and EliteMatrimony : Personalised services for the HNI, NRI and Premium segment of customers Wedding Services: Wedding Bazaar: Largest wedding services provider. : Largest online wedding venues platform. MakeMyWedding: Personalized & curated wedding planning services. New Initiatives & Verticals MatchAstro: It is an exclusive astrology app for marriage, love, and relationship ManyJobs: Employment platform targeting grey collar workforce. In vernacular language for the Tamil Nadu market (initial phase) and with some of industry first features for both companies and candidates : A fin-tech platform to facilitate wedding loans through a curated set of institutions : A next generation serious relationship app. International Presence has successfully expanded into international markets, with UAE operations thriving for 5+ years and Bangladesh operations commencing in 2021. With a mission to dominate the $100 billion wedding services industry.

Apply Now