Madrid, Spain · MSc Completed July 2026 · Available Now

Divyansh
Shrivastava

|
LangGraph · Multi-Agent Systems LLMs · RAG · ChromaDB Python · Databricks · PySpark AWS · GCP · Docker Shipped @ Real Madrid
View My Work LinkedIn
Scroll
01 / About

AI Engineer.
5 years DE.
Ships production systems.

At Real Madrid, I worked as the sole AI Engineer — identified the business problem, designed the architecture, and shipped ClubOS: a multi-agent agentic AI platform deployed on GCP Cloud Run, used by senior club stakeholders for monthly commercial decision-making.

Before Real Madrid, I spent 5 years in production data engineering. At Deloitte, I was Tech Lead for a 4-engineer team — architected an AWS data lakehouse, shipped the team's first RAG system via Amazon Bedrock. At Honeywell, I engineered high-volume pipelines processing 1,000+ source streams.

I don't reach for an LLM when a SQL query or a Python function gives you a deterministic answer. The AI layer sits on top of governed data infrastructure — not instead of it.

MSc Sports Analytics completed July 2026 at Universidad Europea de Madrid (Real Madrid Graduate School). Actively looking for my next role.

Real Madrid C.F.
Sole AI Engineer — shipped ClubOS, a multi-agent agentic AI platform with LangGraph orchestration, ChromaDB vector retrieval, Databricks Medallion lakehouse, deployed on GCP Cloud Run. 604 automated tests.
HashedIn by Deloitte
Tech Lead — architected enterprise AWS data lakehouse on Apache Iceberg, CDC via Kafka, dbt transformation with lineage tracking. Deployed first RAG system to production via Amazon Bedrock.
Honeywell Technology Solutions
Built high-volume production data pipelines processing 1,000+ source streams — ingestion, storage optimization, and query performance at product-company standards.
5+ Years Experience
604 Automated Tests
59 Business Metrics
22 Validated Signals
02 / Experience

Where I've
shipped.

Mar 2026 — Jul 2026
Madrid, Spain
AI Engineer
Real Madrid C.F.
  • Shipped ClubOS — a multi-agent agentic AI platform processing 59 business metrics across 5 digital platforms, benchmarking against 5 elite European clubs, deployed on GCP Cloud Run
  • Built four LangGraph-orchestrated agents: Scout (three-tier hybrid retrieval via metric registry, BM25 skill files, ChromaDB vector search), Watchdog (deterministic anomaly detection), Investigator (ReAct agent with 6 tools), Briefer (executive summary generation)
  • Architected Databricks Medallion lakehouse (Bronze → Silver → Gold) with deterministic priority scoring and 22 statistically validated predictive signals (Pearson r ≥ 0.60) across 103 months
  • Deployed FastAPI backend with 20+ typed endpoints, React/TypeScript frontend, Pydantic v2 schema validation, keyless CI/CD via GitHub Actions — 604 automated tests
  • No-fabricated-numbers guardrail: every AI-generated claim must trace to a source-tagged tool result — ungrounded numbers rejected before reaching stakeholders
PythonLangGraphChromaDBOpenAI APIDatabricksDelta LakeFastAPIReactdbtGCP Cloud RunDocker
Jun 2021 — Oct 2024
Bengaluru, India
Senior Data Engineer
HashedIn by Deloitte
  • Tech Lead for a 4-engineer team — architected AWS data lakehouse on Apache Iceberg with governed multi-zone layout serving 6+ cross-functional teams with column-level security and dbt lineage tracking
  • Deployed RAG pipeline using Amazon Bedrock for internal knowledge Q&A — retrieval architecture, prompt engineering, guardrails; first LLM system shipped to production
  • Built CDC ingestion via Kafka and AWS Glue Streaming — replaced 24-hour batch loads with sub-hour SLA delivery across critical reporting domains
  • Refactored Databricks/PySpark workloads on Delta Lake — reduced job execution time 40%; semantic data layer with star schemas for QuickSight and Power BI
  • Built platform observability (Grafana), shared Python libraries, GitLab CI pipelines — cut onboarding time 50%, reduced production defects 30%
PythonPySparkdbtAirflowKafkaAWSGCPDatabricksDelta LakeIcebergBedrockTerraform
Oct 2019 — May 2021
Bengaluru, India
Big Data Engineer
Honeywell Technology Solutions
  • Built high-volume production data pipelines processing 1,000+ source streams — ingestion from Oracle and MySQL into HDFS with schema validation and deduplication at source
  • Boosted distributed query performance 35% through strategic partitioning, bucketing, and YARN cluster tuning
  • Designed ETL datasets feeding predictive maintenance ML models — enabling early-failure detection and real-time operational monitoring
PythonSQLHadoopHiveSqoopHDFSPySparkOracle DB
03 / Projects

Things I've
built.

Real Madrid 01

ClubOS

Agentic AI Platform

Multi-agent agentic AI platform built as sole AI Engineer at Real Madrid. Four LangGraph-orchestrated agents process digital channel data, benchmark against elite European clubs, and deliver ranked priorities with predictive signals to club leadership. Deployed on GCP Cloud Run.

59Metrics
22Signals
604Tests
103Mo. History
LangGraphChromaDBOpenAI APIDatabricksFastAPIReactdbtGCPDocker
GitHub →
MSc Final Project 02

ScoutIQ

Football Scouting Intelligence Platform

Production-grade scouting platform covering 15,000+ players across Europe's top 5 leagues. Role-based scoring engine across 13 positional profiles, player similarity engine, and AI Scout Assistant powered by the Anthropic API.

15K+Players
13Profiles
5Leagues
20+Metrics
FastAPIPostgreSQLDuckDBAnthropic APIReactAirflowDockerdbt
GitHub →
03

Urban Pulse

Multi-Cloud Data Lakehouse

Production portfolio project ingesting NYC 311, NYPD crime, TfL London transit, AirNow EPA, and Open-Meteo data into a multi-cloud lakehouse across GCP, AWS, and Databricks. Full dbt transformation layer with automated data quality testing.

5APIs
33+Unit Tests
4Airflow DAGs
3Clouds
GCPAWSDatabricksdbtAirflowDockerBigQueryTerraform
GitHub →
04

StatsBomb Viz

Tactical Analysis Dashboard

Interactive football analytics dashboard using StatsBomb open event data. Pass networks, shot maps, player heatmaps, pressing metrics. Key finding: Leicester City's 2015/16 title was built on deliberately low pressing intensity — 3rd lowest PPDA among top-6 clubs.

PythonmplsoccerStreamlitStatsBomb APIPlotlypandas
GitHub →
04 / Skills

My toolkit.

AI & Agentic Engineering
LangGraph · Multi-Agent SystemsProduction
RAG Pipelines · Vector SearchProduction
OpenAI API · Anthropic APIProduction
Prompt Engineering · GuardrailsProduction
ChromaDB · LangSmith · RAGASAdvanced
Data Engineering
Python · SQLExpert
PySpark · DatabricksExpert
dbt · Apache AirflowExpert
Kafka · CDC IngestionAdvanced
Cloud & Infrastructure
AWS (S3, Glue, EMR, Bedrock)Expert
GCP (BigQuery, Cloud Run)Advanced
Docker · Terraform · CI/CDAdvanced
FastAPI · React · TypeScriptAdvanced
Football Analytics
StatsBomb · Wyscout · SkillCornerAdvanced
xG · xA · PPDA · Progressive MetricsAdvanced
Also work with
PostgreSQLDuckDBDelta LakeApache IcebergMLflowPydanticStreamlitPlotlymplsoccerTableauPower BISnowflakeMCP
05 / Education

Academic
foundation.

2025 — 2026
MSc Sports Analytics
Universidad Europea de Madrid
Real Madrid Graduate School
  • AI Engineering internship at Real Madrid — built and shipped ClubOS (agentic AI platform)
  • Final Project: ScoutIQ — 15K+ player scouting intelligence platform
  • Football performance analytics, ML for scouting, statistical modelling
  • Completed July 2026
2019
PG Diploma, Big Data & Analytics
IACSD, Pune
  • Distributed computing & Hadoop ecosystem
  • Data warehousing and analytics engineering
  • Technical foundation for enterprise DE career
2014 — 2018
B.E. Mechanical Engineering
ITM Universe, Gwalior
  • Engineering fundamentals, systems thinking
  • Analytical problem-solving grounding

Let's build
something
together.

MSc completed. 5 years of production engineering. Shipped an agentic AI platform at Real Madrid. Looking for my next role — if you're building AI systems, data platforms, or sports analytics, let's talk.

Email divyyansh99@gmail.com → LinkedIn linkedin.com/in/divyyansh05 → GitHub github.com/divyyansh05 →
Available immediately
What I'm looking for
  • AI Engineer at VC-backed startups
  • Forward Deployed Engineer
  • Senior Data Engineer at product companies
  • Applied AI · ML Engineer roles
  • Football Analytics at clubs & sports data companies