Himanshu Pant — Data Engineer | AI & Infrastructure

About

Engineer at heart.

Data Engineer with expertise in large-scale data infrastructure, real-time processing, and AI/ML systems. I've migrated 1B+ records to distributed architectures, eliminated 90-minute production bottlenecks, and built real-time pipelines handling 500K+ daily events with 99.9% accuracy.

Track record of rapid impact across insurance, sports analytics, and healthcare — promoted within 6 months at Super Six and received SPOT Recognition Award for delivering critical data infrastructure under tight deadlines.

🏥 Healthcare⚽ Sports Analytics🏦 Insurance & Banking📊 Retention & Segmentation

100M-1B

Records Processed

M.S.

Software Engineering, Data Science Minor — ASU

Tech I work with

Languages

PythonSQLJavaScriptPySpark

Data & ML

Apache SparkDatabricksAirflowLangChainSnowflakeGreat Expectations

Cloud & Infra

AWS (EMR, S3, Lambda, SQS, ECR)AzureDockerJenkinsCI/CD

Frontend

ReactNext.jsNode.jsHTML/CSS

Databases

PostgreSQLSQL ServerOracleNoSQLVector DBs

Tools

GitPower BIStreamlitJira

Achievements

Highlights.

🏆

DEVHACKS 2026 — 1st Place

Won Track 1 with MeetFlow — intelligent task orchestration converting meeting transcripts into capacity-aware ticket assignments using LLM-powered analysis, competing against 100+ teams.

📜

Microsoft Certified: Fabric Analytics Engineer

Passed DP-700 (May 2026) — validated expertise in Microsoft Fabric analytics engineering, data warehousing, and cloud data solutions.

🏆

HackASU — FairCharge

Built a medical bill audit pipeline at HackASU that uses Claude Vision + SapBERT to detect overcharges, flagging $1,300+ in average billing errors per hospital bill.

🏅

SPOT Award — Exceptional Delivery

Recognized at Super Six Sports Gaming for exceptional delivery and cross-team collaboration on critical product features.

🎓

M.S. Software Engineering

Arizona State University — Data Science Minor — 2026. Focus on AI/ML infrastructure, distributed systems, and agentic architectures.

Experience

Where I've worked.

AI Engineer

MyEdMaster

US-based EdTech Company (Virginia)

Jan 2026 – Apr 2026 Legal Tech / AI

Remote

Built multi-agent RAG system for personalized legal guidance — created LangGraph orchestration pipeline with Qdrant vector DB, achieving sub-2s query latency across 10K+ legal documents
Designed data ingestion and transformation pipelines converting unstructured legal corpora into structured, queryable vector embeddings with real-time personalization signals via LLM APIs, achieving 95% relevance accuracy
Built FastAPI + Node.js backend integrating vector DB lookups, session state management, and multi-turn conversation history, supporting 100+ concurrent sessions with <200ms response time

PythonLangChainVector DBMulti-Agent SystemsLLMRAG

Consultant II — Analytics (Data Engineer)

EXL

Jul 2023 – Mar 2024 Insurance

Gurugram, India · Remote

Cut PySpark ETL processing time by 66% (90 min → 30 min) on AWS EMR for insurance reporting stakeholders
Built automated data validation with Great Expectations across 5 insurance data sources, catching quality issues before downstream models
Orchestrated end-to-end Airflow workflows integrating APIs, databases, and S3 for underwriting ML models
Implemented CI/CD pipelines (Jenkins + Bitbucket) to streamline releases and reduce deployment friction

PySparkAWS EMRAirflowSnowflakeGreat ExpectationsJenkins

Data Engineer

Super Six Sports Gaming

Aug 2022 – Jul 2023 Sports Analytics

Gurugram, India · On-site

Designed and built the entire data infrastructure from scratch — data models, ingestion pipelines, and processing layers — serving as the foundation for all analytics and ML workloads
Developed and deployed a churn prediction model for new users, directly reducing churn and driving measurable improvements in retention and revenue
Built end-to-end pipelines (Python, PySpark, SQL) ingesting semi-structured data from APIs, logs, and NoSQL sources at scale
Automated data validation, monitoring, and unit testing frameworks, improving quality and reliability across the full pipeline

PythonPySparkSQLNoSQLMachine LearningData Modeling

Associate Data Engineer

Futurense Technologies

Oct 2021 – Jul 2022 Healthcare

Bangalore, India · Remote

Led migration of 50 legacy SAS batch workflows to Spark on Azure Databricks for financial services clients
Cut runtime from 6 hours to under 50 minutes (86% reduction) via partitioning, caching, and broadcast joins
Delivered zero-data-loss transition across all 50 workflows, resolving schema mismatches and validation gaps
Built automated data quality checks and transformation pipelines for downstream ML and reporting

Azure DatabricksApache SparkPySparkSASSQLS3

Executive Analyst

Koron Projects Limited

Oct 2018 – Jul 2021 Construction & Infrastructure

Gurugram, India · On-site

Built 15 Power BI dashboards providing real-time visibility across $50M+ in annual construction projects
Automated monthly reporting, replacing manual Excel workflows and saving 20 hours/month for finance stakeholders
Designed KPI dashboards tracking 50+ active projects, enabling leadership to identify delays and make faster decisions

Power BISQL ServerOraclePythonExcel

Projects

Selected work.

FairChargeHackathon

Solo builder — HackASU Claude AI Builder Hackathon

A medical bill audit pipeline that reads your bill, identifies every charge, benchmarks against real CMS Medicare pricing data for your state, detects overcharges and billing violations, and generates a ready-to-send dispute letter. Built to fight the information asymmetry where 49–80% of medical bills contain errors.

Bill Upload & Analysis

1 / 3

PythonClaude VisionSapBERTChromaDBSQLiteStreamlitCMS Data

Code

MeetFlowHackathon Winner

Led the LLM pipeline & orchestration layer

Intelligent task orchestration that converts meeting transcripts into actionable tickets. Analyzes via GPT-4o-mini, checks team capacity through Taiga, recommends smart reassignment for overloaded members, and notifies via Slack.

Agent 1: Transcript Analysis

1 / 3

PythonStreamlitOpenAITaiga APISlack

Code

Serverless Face RecognitionAcademic

A serverless, multistage face recognition system using edge computing. IoT clients send video frames processed through decoupled detection and recognition stages via event-driven architecture — scalable, real-time identification without persistent servers.

AWS LambdaSQSECRDockerPyTorchOpenCVEdge Computing

LifeSyncLive

Full-stack productivity platform with Google OAuth, Focus Score tracking, daily task management, Pomodoro timer, and motivational micro-challenges.

Focus Score Dashboard

1 / 2

ReactNode.jsGoogle OAuthREST API

Live

The Pause ButtonLive

A developer wellness app with mixable ambient soundscapes, 10+ casual mini-games, and a calming UI — because even builders need a break.

Nature Sounds Mixer

1 / 2

ReactJavaScriptWeb Audio APICSS Animations

Live

Legal AI — Agentic RAGBuilding

Multi-agent system providing accessible legal guidance for those who can't afford a lawyer. Extensive query refinement via vector DB, specialized agents that collaborate independently.

PythonLangChainVector DBMulti-AgentLLM

Feature Store with Time TravelIndustry

Centralized Feature Store using Apache Iceberg with time-travel, enabling ML teams to train, test, and deploy from a single source of truth.

Apache IcebergSparkPythonML Infra

SAS to Spark MigrationIndustry

Migrated 50 legacy SAS workflows to Spark, cutting runtime from 6+ hours to under 50 minutes via partitioning, broadcast joins, and caching.

Apache SparkPySparkSASData Engineering

I build things
that scale.