AAAbhinav Anand

Hello, I’m

Abhinav Anand

Principal Software Engineer · Identity, Access & Data Platforms · Distributed Java Systems at Scale

I design and build the platforms that decide who gets to see what — identity, access and data governance at enterprise scale, with real users, real security requirements and real production pagers. Right now that means leading AccessIntel, the control plane for Walmart’s centralised authorization platform for HR data (15+ business-critical consuming applications, 3 production regions). Before that I architected Sage, Walmart’s governed agentic-analytics platform, and shipped it to production as the sole engineer in about five months.

I wrote my first BASIC programs long before it was a job, went professional with Java in 2004, ran my own product company for five years, and have spent the last decade as an architect at Cognizant (Apple), Applied Materials and Walmart Global Tech. These days I write the sentence that writes the code — and then review every line of it.

Read the résuméDownload PDFWatch the story →
22 yrs
professional engineering
5 mo
Sage: zero to production
4M+
SSO customers served
50+
mobile apps published
90+
architecture decision records

Selected work

All projects →
Walmart Global Tech · Jul 2026 – present

AccessIntel — HR data authorization control plane

Technical lead for the control plane of Walmart’s centralised authorization platform for HR and people data — deciding who may see or act on employee data for 15+ business-critical applications across 3 production regions. Took it from zero to staging in about five weeks; production rollout in progress.

  • Cut access-anomaly detection from a two-hour batch process to under a minute, with alerting built in.
  • Designed to be agent-operable from day one: AI agents work through the same authorization and audit controls as the human-facing portal.
Java 21Spring BootOPA/RegoPingFederate OIDCKafkaCosmos DB
Walmart Global Tech · Jun 2025 – Jul 2026

Sage — governed agentic analytics

Walmart’s first governed agentic-analytics platform (sage.walmart.com): AI agents turn natural language into dashboards that respect every access control. Architected and shipped to production as the sole engineer — first release in ~5 months, 60+ architecture decision records.

  • Every query — human- or agent-authored — is checked and rewritten for the viewer before it touches data, which is what makes letting AI write analytics safe.
  • Agents use the same governed interface as people: certified queries exposed as MCP tools, nothing more.
Java 21Spring BootPythonFastAPIsqlglotReact 19
Walmart Global Tech · ULearn International · 2025

HackerOne bug-bounty remediation

An unplanned security programme: 23 in-scope HackerOne + Snyk findings (15 Critical, 8 High) across 7 microservices, closed under a 3-week SLA with zero production incidents.

  • Re-architected the JWT model: single mint point, server-side RBAC, no trust in client-supplied parameters.
  • Stood up an 882-assertion SQL-injection regression suite, retained as a permanent CI gate.
JavaSpring SecurityJWTSnykCI gates
Walmart Global Tech · 2022 – 2025

OIDC SSO for Walmart Mexico (DCP)

OIDC-compliant single sign-on for the Digital Connections Program across Walmart International markets — serving 4M+ customers at an 82% login rate.

  • Identity, profile and membership initiatives across Mexico, Canada and Chile.
  • Chaired the Operational and Engineering Excellence Council — a weekly DORA-style scorecard across 20+ services — and the daily Change Council.
OIDCOAuth2PingFederateSpring BootKubernetes

Now — September 2026

  • Technical lead of AccessIntel at Walmart Global Tech — the control plane for Walmart’s centralised HR-data authorization platform (Java 21 policy decision point, OPA/Rego, Kafka real-time detection, Spring AI MCP server); production rollout in progress.
  • Sage, the governed agentic-analytics platform I built as sole engineer, runs in production at sage.walmart.com (handed over Jul 2026).
  • QuizWrap, my free AI study platform, is at ~420 users across 60+ countries with zero paid acquisition.
  • Open to conversations about principal / staff / architect roles in identity & access, data platforms and AI infrastructure.

More about how I work →

Story · 13 chapters

From GOTO to agents.

A lightly exaggerated true story of the career — from the first BASIC program to directing coding agents — told as an auto-playing deck.

Play the story →

New · playable

Digger — the 1983 classic, rebuilt.

An HTML5 tribute built in one session with Claude Code, with an AI watch mode that plays the game itself — and the story of how it was made.

Play Digger →