TechSci Research - Lead Data Scientist - LLM Models

About the Role :

The design and development of a cutting-edge application powered by large language models (LLMs).

This tool will provide market analysis and generate high-quality, data-driven periodic insights.

You will play a critical role in building a scalable and intelligent system that integrates structured data, NLP capabilities, and domain-specific knowledge to produce analyst-grade content.

Key Responsibilities :

- Design and develop LLM-based systems for automated market analysis.

- Build data pipelines to ingest, clean, and structure data from multiple sources (e.g., market feeds, news articles, technical reports, internal datasets).

- Fine-tune or prompt-engineer LLMs (e.g., GPT-4.5, Llama, Mistral) to generate concise, insightful reports.

- Collaborate closely with domain experts to integrate industry-specific context and validation into model outputs.

- Implement robust evaluation metrics and monitoring systems to ensure quality, relevance, and accuracy of generated insights.

- Develop and maintain APIs and/or user interfaces to enable analysts or clients to interact with the LLM system.

- Stay up to date with advancements in the GenAI ecosystem and recommend relevant improvements or integrations.

- Participate in code reviews, experimentation pipelines, and collaborative research discussions.

Qualifications Required :

- Strong fundamentals in machine learning, deep learning, and natural language processing (NLP).

- Proficiency in Python, with hands-on experience using libraries such as NumPy, Pandas, and Matplotlib/Seaborn for data analysis and visualization.

- Experience developing applications using LLMs (both closedand open-source models).

- Familiarity with frameworks like Hugging Face Transformers, LangChain, LlamaIndex, etc.

- Experience building ML models (e.g., Random Forest, XGBoost, LightGBM, SVMs), along with familiarity in training and validating models.

- Practical understanding of deep learning frameworks: TensorFlow or PyTorch.

- Knowledge of prompt engineering, Retrieval-Augmented Generation (RAG), and LLM evaluation strategies.

- Experience working with REST APIs, data ingestion pipelines, and automation workflows.

- Strong analytical thinking, problem-solving skills, and the ability to convert complex technical work into business-relevant insights.

Preferred :

- Familiarity with the chemical or energy industry, or prior experience in market research/analyst workflows.

- Exposure to frameworks such as OpenAI Agentic SDK, CrewAI, AutoGen, SmolAgent, etc.

- Experience deploying ML/LLM solutions to production environments (Docker, CI/CD).

- Hands-on experience with vector databases such as FAISS, Weaviate, Pinecone, or ChromaDB.

- Experience with dashboarding tools and visualization libraries (e.g., Streamlit, Plotly, Dash, or Tableau).

- Exposure to cloud platforms (AWS, GCP, or Azure), including usage of GPU instances and model hosting services.