AI/Data Engineer focused on building production-grade LLM systems, agentic pipelines, and scalable data infrastructure. Currently at TELUS Digital AI, previously at Tredence Inc.
📍 India | 📧 [email protected] | LinkedIn
- LangGraph-based agentic pipelines for large-scale conversational audio analysis (400K+ hours, deployed to production)
- RAG pipelines with PGVector, Pinecone, and FAISS for semantic search over unstructured data
- Distributed systems — horizontal scaling with Postgres advisory locks, orchestrated via Kubernetes
- Human-in-the-loop data pipelines for LLM fine-tuning (SFT, RLHF) across audio, vision, and language
Solutions Engineer — TELUS Digital AI (Playment) (Nov 2024 – Present) LangGraph agents · RAG · LangChain · OpenAI SDK · Kubernetes · PGVector · Prompt Engineering · RLHF · SFT
Software Engineer — Tredence Inc. (Jun 2023 – Nov 2024) PySpark · Azure Databricks · Azure Data Factory · Terraform · Docker · Data Quality & Observability
SDE Intern — Samsung SDS (Mar 2023 – Jun 2023) Java Spring Boot · React · MySQL · Agile · cleared Samsung Advanced Software Test (DSA)
| Area | Tools |
|---|---|
| AI / LLM | LangChain, LangGraph, OpenAI SDK, RAG, Vector Search, RLHF, SFT |
| Data | PySpark, SQL, Pandas, NumPy, Databricks, MLflow |
| Databases | PostgreSQL, PGVector, Pinecone, FAISS, Weaviate, MySQL |
| Cloud & DevOps | Azure, AWS, Kubernetes, Docker, Terraform, Airflow, Git |
| ML | PyTorch, scikit-learn, OpenCV |
| Languages | Python, C/C++, JavaScript |
Micrograd — Neural Network from Scratch Built a scalar-valued autograd engine with reverse-mode automatic differentiation from first principles; trained an MLP on MNIST.
Game Playing AIs Implemented Minimax, A* search, Q-Learning, and backtracking agents across multiple game environments.
- B.E. Computer Engineering — Thapar Institute of Engineering & Technology (2019–2023) · CGPA: 8.29
- Databricks Certified Data Engineer Associate — Issued Oct 2023
Always open to interesting problems in AI infrastructure, agentic systems, and data engineering. Let's connect!