Vansh BData Engineer

Prepared for GridBase version

Ingestion thatsurvivesschema drift.

Sports-data pipelines at league scale: multi-provider ingestion, schema drift, and analytical stores that answer analyst questions in milliseconds.

6B+

rows ingested into ClickHouse

500ms

P90 on analyst-facing queries

750+

matches auto-ingested per cycle

About me

I am Vansh Bordia, a data engineer who likes the boring parts: idempotent jobs, validated schemas, and a pipeline that can be rerun at 2am without anyone sweating. Half a terabyte of raw broadcast data became structured analytics because the ingestion was built to survive messy sources.

I care about the boring parts because they are what makes the numbers trustworthy downstream: reconciliation checks, retry and rate-limit logic, columnar storage tuned to the actual query pattern, and documentation so the next person can extend it.

Why me

I build the pipes, not just the queries: multi-source ETL that reconciles inconsistent schemas, analytical storage tuned to the query pattern, and ingestion that has not lost a row in production.

I build the pipes, not just the queries. Ingesting multi-source broadcast JSONL with inconsistent schemas, modelling it for analytical workloads, and keeping the pipeline idempotent so a rerun never corrupts a dashboard. 6B+ rows in production ClickHouse at sub-500ms P90.

Multi-source ETL reconciling inconsistent provider schemas

Columnar modelling tuned to real query patterns, not theory

Schema validation and reconciliation before anything is published

Role stack

Python / SQL / ClickHouse / PostgreSQL / Apache Spark / Kafka / ETL Orchestration / Data Modeling / Schema Validation / Query Optimization / Data Quality / Docker

Relevant work

RiftWatch

Esports intelligence platform, acquired by a pro org

  • 750+ pro matches processed into 6B+ analytical rows in ClickHouse
  • 20 orgs at peak including Tier-1 across Americas and EMEA
  • VALORANT-specific Gemma models plus RAG over live patch data
  • Agentic harness routing tools, models and queries; sub-500ms P90
PythonFastAPIClickHousePostgreSQLRedisKubernetesReact
Visit project

VLR Dev API

Type-safe Python library for Valorant data, published on PyPI

  • Type-safe Python library for Valorant data, published on PyPI
  • Retry logic, rate limiting and schema validation built in
  • Earned official Riot data API access in 2 weeks, not 9 months
PythonPyPIREST
Visit project

DDL Showcase

Interactive SQL schema visualizer, 100% client-side

  • Dual-dialect SQL parser, PostgreSQL plus ClickHouse
  • Interactive ER diagrams with auto-layout and relationship inference
  • 100% client-side, data never leaves the browser
TypeScriptReactSQL
Visit project

ACF Import/Export

Production WordPress plugin in the official directory

  • Production WordPress plugin in the official directory
  • Secure file handling and sanitization per WP standards
PHPMySQLWordPress
Visit project

Contact

Building sports data infrastructure? Talk to me.

GitHubLinkedInBehanceDribbbleKaggleInstagramTwitterFacebookYouTubeTwitchSpotifyReddit