Bhavyadeep Singh Sikarwar

AI Data Engineer

Indore, India 2+ yrs exp 90 · Outstanding

About

AI Data Engineer who takes full ownership of what I build. At Cubexo, I independently designed and shipped production-grade data pipelines for an international real estate project — end-to-end, from web scraping through ingestion, transformation, and storage into PostgreSQL. Alongside that, I built a Hybrid RAG chatbot with multi-hop retrieval and guardrails using LangChain and LangGraph. I’m someone who learns fast from projects and people around me, takes responsibility without being pushed, and actively looks for hard problems to solve.

Skills & Expertise (42)

Apache Airflow Advanced

9.0/10

Years Exp

SQL Advanced

9.0/10

Years Exp

Python Advanced

9.0/10

Years Exp

NumPy Advanced

8.5/10

Years Exp

Pandas Advanced

8.5/10

Years Exp

ETL Design Advanced

8.5/10

Years Exp

Pipeline Orchestration Advanced

8.5/10

Years Exp

RAG Advanced

8.5/10

Years Exp

PySpark Advanced

8.0/10

Years Exp

Docker Compose Advanced

8.0/10

Years Exp

Docker Advanced

8.0/10

Years Exp

FastAPI Advanced

8.0/10

Years Exp

LangGraph Advanced

8.0/10

Years Exp

LangChain Advanced

8.0/10

Years Exp

BM25 Advanced

8.0/10

Years Exp

Snowflake Advanced

8.0/10

Years Exp

Vertex AI Advanced

8.0/10

Years Exp

Cloud Storage Advanced

8.0/10

Years Exp

BigQuery Advanced

8.0/10

Years Exp

Apache Spark Advanced

8.0/10

Years Exp

Cloud Composer Advanced

8.0/10

Years Exp

Postgresql Intermediate

7.5/10

Years Exp

MySql Intermediate

7.5/10

Years Exp

SQL Server Intermediate

7.5/10

Years Exp

Power BI Intermediate

7.5/10

Years Exp

Data Warehousing Intermediate

7.5/10

Years Exp

ChromaDB Intermediate

7.0/10

Years Exp

Hypothesis Testing Intermediate

7.0/10

Years Exp

EDA Intermediate

7.0/10

Years Exp

Apache Kafka Intermediate

7.0/10

Years Exp

Java Intermediate

7.0/10

Years Exp

Dataproc Intermediate

7.0/10

Years Exp

Langsmith Intermediate

7.0/10

Years Exp

Groq Intermediate

7.0/10

Years Exp

Gemini API Intermediate

7.0/10

Years Exp

Google ADK Intermediate

7.0/10

Years Exp

Web Scraping Intermediate

6.0/10

Years Exp

Hive Intermediate

6.0/10

Years Exp

Hadoop Intermediate

6.0/10

Years Exp

Pub Sub Advanced Excel

Work Experience

Data Engineer

Cubexo Solution

Apr 2025 - Present

Designed and owned production-grade ETL pipelines end-to-end, as measured by 50K–500K records processed per weekly cycle with zero manual intervention, by building Python + SQL ingestion, normalization, and scheduling workflows using Apache Airflow. Engineered pipeline resilience under a hard geo-restriction constraint, as measured by uninterrupted client-side operation, by architecting Airflow DAGs that routed execution around infrastructure limitations without human touch. Improved data quality and downstream accuracy, as measured by 90% backcasting accuracy against 2010 historical actuals, by identifying and resolving 20x price outliers and reconfiguring asset-type base periods. Delivered analytics-ready data powering 5 live market BI indexes (Price, Volume, Auction, Affordability, Volatility), signed off by senior leadership and published publicly for self-serve market intelligence.

Hybrid RAG Chatbot Developer

Cubexo Solution

Apr 2025 - Present

Architected a production-grade Hybrid RAG chatbot with multi-hop retrieval, as measured by accurate resolution of queries requiring 2+ chained document lookups, by combining BM25 sparse retrieval with dense vector search and Reciprocal Rank Fusion (RRF). Implemented input/output guardrails to prevent hallucination and out-of-scope responses, as measured by consistent query rejection on irrelevant/unsafe inputs, by wrapping the FastAPI inference endpoint with validation layers. Built a structured offline + online evaluation harness, as measured by tracked retrieval precision and answer relevance scores pre/post RRF tuning, using LangSmith for tracing and a curated real-estate query benchmark set. Containerized the full inference stack, as measured by single-command reproducible deployment, using Docker Compose to orchestrate FastAPI, ChromaDB, and the LLM layer.