Lead Data Engineer · Data + AI · Cloud & Big Data

Engineering the data foundations behind smarter businesses.

Biography

Pallaav Sethi — a Lead Data Engineer building scalable data platforms, pipelines and AI-powered solutions across AWS, Azure and the modern data ecosystem.

I design the architecture and then build it: cloud data platforms, production pipelines and intelligent systems that turn raw data into decisions businesses can trust.

Portrait of Pallaav Sethi, Lead Data Engineer and freelance data consultant

Years of experience

0+

NUMBER OF PROJECTS DELIVERED

0

Platforms I build with

AWS logo
Databricks logo
MongoDB logo
Google logo
Python logo
Snowflake logo
Microsoft Azure logo
Apache Spark logo
Apache Airflow logo
Apache Kafka logo

Impact

Numbers from real delivery

0+

Years of experience

0%

Pipeline processing time reduced

0%

Cloud & data cost reduced

0%

Reduction in data errors

0%

MoM sales performance uplift

About

Engineering systems that scale.

I work where architecture, engineering and AI meet — across financial services, enterprise data platforms, fraud and risk management, cloud modernisation, governance and data warehousing.

My work spans the full lifecycle of a data platform: understanding the business problem, designing the architecture, engineering the pipelines, running it on cloud, layering AI on top and then optimising for performance and cost. I have built fraud detection pipelines, enterprise finance foundations, credit and bureau data marts and an AI-powered sales assistant running in production.

B.Tech, Computer Science & Engineering — SRM Institute of Science and Technology, Chennai (2021), with a minors in AI/ML and Data.

Data platform blueprint

From raw signals to decisions

Sources

Pipelines

Cloud

Analytics

AI

Reliable
Governed
Scalable

Expertise

A technology ecosystem, not a checklist.

Data Engineering

PythonSQLPySparkApache SparkETL / ELTData PipelinesData ModellingData Warehousing

Cloud

AWSAzureADLSS3LambdaEMRGlueRedshiftKinesisAthenaStep FunctionsAPI Gateway

Modern Data Stack

DatabricksSnowflakeDelta LakeApache IcebergMicrosoft FabricUnity Catalog

Streaming & Distributed

KafkaSpark StreamingChange Data CaptureEvent-drivenReal-timeBatch

Data Architecture

Data LakeLakehouseMedallionDistributed SystemsSystem DesignLow-latency

AI & Intelligent Systems

Generative AIAI-powered applicationsVector DatabasesIntelligent AutomationBehavioural Analytics

Services

What I Can Build For You

01

Modern Data Platforms

Data lake and lakehouse platforms on cloud, with warehouses, modelling and ETL/ELT designed for both batch and streaming workloads.

Data lake / lakehouseCloud data platformsWarehousing & modellingBatch + streaming
Build My Data Platform →
02

Data Pipeline Engineering

Reliable ETL/ELT pipelines that integrate APIs and enterprise sources, with quality checks and performance tuning built in.

API & source integrationStreaming and CDCData quality checksPipeline optimisation
Build My Data Platform →
03

Data + AI Solutions

AI-powered internal tools, assistants and automated workflows grounded in your own data — like ARYA, an AI sales assistant I built in production.

AI assistants & internal toolsData-driven automationVector databasesAI/data infrastructure
Build My Data Platform →
04

Cloud & Data Modernisation

AWS and Azure architecture, migration off legacy systems, and performance work on the infrastructure your data runs on.

AWS / Azure architectureMigration & modernisationPerformance engineeringCloud infrastructure
Build My Data Platform →
05

Data Quality & Governance

Governance frameworks, validation rules and data models that make reporting trustworthy and audits straightforward.

Governance frameworksValidation & qualityRobust data modelsCompliance controls
Build My Data Platform →
06

Data Platform Cost Optimisation

Reduce platform spend without sacrificing performance — a 60% reduction across AWS and Snowflake on a live enterprise estate.

Spend analysisWorkload right-sizingQuery & storage tuningSustained savings
Build My Data Platform →

Experience

Where I have built.

Sep 2025 — Present

BP logo

Senior Data Engineer — BP

Building fraud and risk data pipelines across Azure Data Factory, Databricks and ADLS, automating fraud detection with Auto Loader-based ingestion and contributing to the Finance Data Foundation and wider cloud modernisation.

Jul 2022 — Aug 2025

Piramal Finance logo

Data Engineer — Piramal Finance

Delivered Credit and Bureau DataMarts on Snowflake from 10+ source systems with governance controls, built AWS Glue pipelines and the Credit Scrub process, drove a 40% cloud cost reduction and engineered ARYA, an AI-powered sales assistant with behavioural analytics.

Nov 2021 — Jun 2022

Toyota Connected logo

Software Engineer — Toyota Connected

Built real-time ingestion and analytics for vehicle performance data using Java microservices, Lambda, DynamoDB and API Gateway, with infrastructure as code and CI/CD pipelines.

Jun 2021 — Oct 2021

EY logo

Technology Consultant — EY

Delivered IT audits, information security and DevOps compliance reviews, automating risk assessment work in Python alongside business stakeholders.

Selected Work

Selected Case Studies

Financial Services

Fraud Risk Data Platform

Problem
Fraud and high-risk transactions were being spotted too late across large, fragmented transaction datasets.
Architecture
Cloud ingestion into ADLS with Databricks processing and an incremental Auto Loader pattern feeding risk models.
Engineering
Automated pipelines, anomaly-detection logic and quality gates across large-scale daily transaction volumes.
Outcome
Faster identification of fraudulent, anomalous and high-risk transactions.
Discuss a Similar Challenge →

Enterprise Finance

Finance Data Foundation

Problem
Finance reporting depended on scattered SAP and enterprise sources with no single trusted layer.
Architecture
SAP & enterprise sources → ADF → ADLS → Databricks → Silver → Gold → Analytics.
Engineering
Medallion modelling, curated Gold datasets and governed transformations serving analytics teams.
Outcome
One consistent, governed finance layer powering downstream reporting.
Discuss a Similar Challenge →

Data + AI

AI Sales Assistant (ARYA)

Problem
Field sales teams lacked timely, data-driven guidance on which customers to prioritise.
Architecture
An AI-powered assistant layered on behavioural analytics and warehouse data, delivered as a production product.
Engineering
Data pipelines, behavioural feature engineering and AI-driven recommendations wired into daily sales workflows.
Outcome
13–15% month-on-month improvement in sales performance.
Discuss a Similar Challenge →

Optimisation

Data Platform Optimisation

Problem
Long-running pipelines and rising cloud spend were slowing delivery and inflating cost.
Architecture
Re-engineered Spark and Snowflake workloads with better partitioning, scheduling and storage design.
Engineering
Pipeline tuning, warehouse right-sizing and query optimisation across the AWS and Snowflake estate.
Outcome
60% faster pipeline processing, 40% lower cloud and data cost, 85% fewer data errors.
Discuss a Similar Challenge →

How I Work

Four steps, no mystery.

01

Discover

Understand the business problem, the existing architecture and the real state of your data landscape.

02

Architect

Design a scalable, cost-conscious solution with a clear delivery path and trade-offs made explicit.

03

Build

Engineer production-ready pipelines, platforms and AI/data systems, hands-on and in reviewable increments.

04

Optimise

Measure performance, reliability, data quality and business impact — then tighten what matters.

Certifications

Credentials behind the work.

AWS Certified logoFeatured

Solutions Architect – Associate

AWS Certified

Databricks logoFeatured

Generative AI

Databricks

MongoDB logo

Data Engineer – Associate

MongoDB

Google logo

Data Analytics Professional Certificate

Google

University of Michigan logo

Python Data Structures

University of Michigan

McKinsey & Company logo

Forward Deployed Program

McKinsey & Company

Digital Privacy India logo

DPDP for Engineers

Digital Privacy India

Have a data problem worth solving?

Whether you are building your first data platform, modernising an ageing stack or turning your data into AI products, I can help you design it and build it.

Contact

Let's build something useful.

Tell me about your project and goals

01
02
03
04
05
06
07