Principal Engineer Compiler Salem (India)

The Prospect

We are building a cutting‑edge AI platform to automate complex enterprise technology transformations. We are moving beyond straightforward "prompt engineering" to build robust, industrial‑grade systems that understand code structure as well as human language.
We are looking for a Language Engineer who understands that reliable code generation requires deep structural understanding, not just large language models. You will be a technical lead of a high‑visibility product team with the autonomy to define technical direction.

The Challenge

You will build the core engine responsible for ingesting, parsing, and analyzing millioms of lines of legacy code (SQL, Teradata BTEQ, SAS etc.). Your work will bridge the gap between Static Analysis and Generative AI, ensuring our models operate with mathematical precision.

What You Will Do

• Deep Code Parsing: Design and implement robust parsers to convert legacy codebases into structured data representations (using tools like ANTLR or similar libraries).
• Static Analysis: Build algorithms to trace data lineage, identify dependencies, and extract logic from unstructured code.
• AI Integration: Collaborate with AI engineers to feed high‑fidelity structural context into our generative models.
• Performance Engineering: Optimize the processing of massive code repositories to run efficiently at scale.

What We Need

• Core CS Fundamentals: Deep understanding of Data Structures, Algorithms, and Graph Theory.
• Language Theory: Experience with Abstract Syntax Trees (ASTs), Lexers, and Parsers is non‑negotiable. You understand why Regex is insufficient for code analysis.
• Engineering Excellence: Solid proficiency in Python and experience building scalable backend systems.
• Problem Solving: You love reverse‑engineering how languages work and finding patterns in chaos.

Preferred Experience

• Experience with ANTLR, Tree‑sitter, or SQLGlot.
• Background in building transpilers, linters, or code migration tools.
• Familiarity with Vector Databases or LLM orchestration frameworks (Lang Chain/Llama Index) is a plus.