LLM Engineer / Generative AI Engineer

About vunaivunai is building AI-powered simulations of real-world populations by creating digital twins of individuals that can replicate their behaviors, preferences, opinions, and decision-making patterns.

Our goal is to simulate a meaningful subset of the Indian population and enable businesses to understand how real people might think, respond, and make decisions in different scenarios.

We're looking for an LLM Engineer / Generative AI Engineer to help build the intelligence layer powering these digital twins.

What You'll Build

• Human Digital Twins: Develop AI systems that model individual personalities, preferences, beliefs, and behavioral patterns using real-world data.
• LLM Architectures: Design and implement LLM-powered systems with structured memory, retrieval, personalization, and context management.
• Agentic Systems: Build AI agents capable of reasoning, maintaining consistent identities, interacting with simulated environments, and responding to changing scenarios.
• Behavioral Simulation: Experiment with techniques to make AI-generated responses more representative of individual human behavior rather than generic LLM outputs.
• Evaluation Systems: Develop methods to measure behavioral consistency, response quality, and alignment between simulated and observed human responses.
• Scalable Intelligence: Help evolve individual digital twins into systems capable of simulating large, diverse population segments.

What We're Looking For

• Strong hands-on experience building applications and systems using LLMs and Generative AI.
• Proficiency in Python and experience with frameworks such as PyTorch, Hugging Face Transformers, or equivalent tools.
• Experience with RAG, embeddings, vector databases, structured memory, and context retrieval.
• Understanding of LLM fine-tuning, supervised fine-tuning, LoRA/QLoRA, or other model adaptation techniques.
• Experience designing evaluation pipelines for LLM outputs, including systematic testing and benchmarking.
• Ability to reason about system architecture, make technical trade-offs, and justify decisions.
• A builder's mindset: comfortable experimenting, prototyping, measuring results, and iterating quickly.

Good to Have

• Experience with multi-agent systems, agent orchestration, or simulation environments.
• Knowledge of synthetic data generation, behavioral modeling, or computational social science.
• Experience with preference learning, reinforcement learning, or model alignment.
• Familiarity with distributed inference, model serving, and optimizing latency and cost at scale.

Who Should Apply?We're interested in engineers who go beyond prompt engineering and understand how to build reliable AI systems. You might be a strong fit if you've built an LLM application from scratch, fine-tuned a model, designed an evaluation framework, or developed an agent that maintains meaningful state across interactions.

We value technical depth, experimentation, ownership, and the ability to defend your architectural decisions over years of experience or impressive-sounding certifications.

Role Details

• Location: Surat, Gujarat
• Work mode: On-site (We are currently not open to remote settings)
• Employment: Full-time
• Experience: Open to strong early-career candidates and experienced engineers with demonstrable hands-on work

ApplicationApply with your resume and links to relevant GitHub repositories, technical projects, or research.

Please highlight one AI/LLM system you've built, the technical challenges you faced, and the decisions you made to solve them.

At vunai, you'll work on a challenging problem at the intersection of Generative AI, human behavior, and population-scale simulation.