Generative AI Architect (Code Generation)

About the role

This position is for a highly capable Generative AI Architect who can design and deliver advanced AI systems for code generation, code optimisation, and automation. The work centers on building complete intelligent solutions using large language models, retrieval-based pipelines, and scalable backend systems to improve developer productivity and modernize software engineering workflows.

Core responsibilities

You will be expected to architect, build, and deploy LLM-driven applications that support code creation, completion, and translation. The role also involves mentoring junior engineers, setting technical direction, and leading projects from proof of concept through production release.

You will need to create evaluation frameworks that verify generated code is accurate, safe, and suitable for real-world use.

Technical requirements

The role calls for strong Python expertise along with solid command of one or more additional languages such as C, C++, Java, or JavaScript. A deep grasp of data structures and algorithms, object-oriented programming, transformer models, tokenization, embeddings, and code-focused LLMs such as StarCoder and CodeLlama is essential.

Preferred experience includes building and maintaining large code datasets through cleaning, deduplication, and structuring, as well as working with Hugging Face Transformers, LangChain, LlamaIndex, and vector databases.

Hands-on experience with retrieval-augmented generation pipelines, including chunking, retrieval, and context optimization, is important. You should also be comfortable with API design, microservices, scalable model deployment, test generation, benchmarking, and security validation.

Familiarity with Docker, Kubernetes, cloud environments such as AWS, GCP, or Azure, and GPU orchestration will be useful.

Mindset and working style

The organization is looking for someone who thinks long term, works with a founder’s mindset, and can turn a vision into a functioning system. Strong ownership, good judgment under uncertainty, comfort with early-stage ambiguity, and a bias toward action are all important.

The ideal candidate should be able to take responsibility across product, architecture, and execution, and should be motivated by building something meaningful rather than merely functional.

Compensation and benefits

The compensation offered is INR 12,00,000 to 24,00,000 per year. The opportunity also includes a co-founder-level role with strategic decision-making authority, equity of up to 5% depending on contribution and milestones, and the chance to work with full ownership rather than in a conventional employee setup.

Additional benefits include the possibility of being named as a co-inventor on patents, direct influence over the product vision, architecture, and roadmap, and the opportunity to help build a category-defining product named SPROSE. The work environment is designed to offer flexibility, high autonomy, and strong accountability.

Eligibility

Applicants must have at least 6 years of experience. Candidates from Computer Science Engineering backgrounds are eligible to apply.

Additional details

This is a full-time onsite role based in Bangalore, Karnataka, India. There is 1 opening available, and the selected candidate is expected to start immediately. The application deadline is 2026-10-21 23:59:59.