Release Train Engineer · AI Product Builder

NishantPancholi

The delivery leader who ships().

15+ years running Agile programs. Today: a 5-team, 52-person Release Train behind a statewide Election Management System — and a live SaaS product built solo with Claude Code.

claude — session — nishantpancholi.com

Terminal transcript: a Claude Code session reads the résumé, coordinates content-strategist, art-director, engineer, and reviewer agents, writes this page with zero dependencies, and ships it in one session.

15+

years leading Agile programs

52

practitioners, one Release Train

20+

retro formats adopted enterprise-wide

9

AI agents in my toolkit

01 — About

Two disciplines, one operator.

For 15+ years I've led Agile delivery where "move fast" meets "heavily regulated": government, mortgage finance, healthcare, retail. Today I run a 5-team, 52-person Agile Release Train delivering a statewide Election Management System — PI Planning, release readiness, cross-team dependencies, budgets, vendors, executive reporting. I also build — a live SaaS product, AI delivery tooling leadership uses daily, and this site. Shipping, it turns out, is a transferable habit.

The program side

Runs the train, keeps it honest

PI Planning, ART Sync, Scrum of Scrums, release readiness, dependencies, and impediment escalation for a system that cannot quietly fail. If it isn't measured, it's a rumor.

  • 5 teams · 52 practitioners on one Release Train
  • Budgets, vendors, and RAID owned end to end
  • Monthly Business Reviews executives actually read
  • Mentor to 4 Delivery Leads and multiple Scrum Masters

The builder side

Builds the tools himself

When the metrics I needed didn't exist, I built the tool that produces them — an AI-assisted dashboard reading live Azure DevOps and BambooHR data. Then I kept going until the tooling became products.

  • Discrepancy Dashboard — the delivery numbers leadership trusts
  • Kensho Cards — live SaaS with real users, built solo
  • 9-agent, 28-skill Claude Code delivery toolkit

02 — Experience

The track record.

  • Leading 5 engineering teams — 52 practitioners — across parallel workstreams including state certification and legacy-system data migration, from PI Planning through release readiness.
  • Owning the full ART cadence — PI Planning, ART Sync, Scrum of Scrums, system demos, and release-readiness reviews — and serving as the single escalation point for cross-team dependencies and impediments.
  • Tracking and managing delivery budgets and resource allocation across the train; coordinating vendor relationships and contract deliverables against program milestones.
  • Designed and built an AI-assisted Discrepancy Dashboard on live Azure DevOps and BambooHR data — 11 automated backlog-quality rules, daily hours and capacity tracking, committed-vs-injected scope trends, capacity forecasting, auto-generated release notes — now leadership's single source of delivery truth.
  • Running executive-level reporting: program roadmaps, milestone plans, RAID logs, sprint demos, release notes, and Monthly Business Review presentations.
  • Mentoring 4 Delivery Leads and multiple Scrum Masters; promoted from Agile Delivery Lead into the RTE seat on the strength of delivery performance.
  • Led delivery for 2 Agile teams in the Single-Family organization, preparing loan-level mortgage data for external vendor analytics and risk-based loan pooling — the data secondary-market investors use to evaluate and purchase loans by risk profile.
  • Owned project plans and delivery timelines across both teams: milestones, data-delivery commitments to vendors, and the handling of sensitive borrower data inside a regulated GSE environment.
  • Managed cross-team risks and dependencies; ran recurring status reporting to leadership covering progress, risk posture, and delivery forecasts.
  • Coached teams through Scrum adoption; mentored Scrum Masters and Product Owners on estimation, story decomposition, and metrics-driven continuous improvement.
  • Led Agile delivery for credit policy technology teams at a Fortune 100 bank — project planning, milestone tracking, risk escalation, and predictable sprint execution in a heavily regulated environment.
  • Created 20+ themed retrospective formats adopted across the enterprise Agile community, improving remote collaboration, onboarding, and inspect-and-adapt behavior.
  • Served as PI Demo champion, coaching teams to communicate released product value directly to business stakeholders.
  • Led Kanban-to-Scrum transitions where workstreams required them.
The AvengersThe WitcherShark TankRockyI am Groot!Jedi Masterbrowse the collection ↗
  • Supported 7 Agile teams modernizing the CMS Health Plan Management System (HPMS) — the federal platform used to administer Medicare health and drug plan data.
  • Coordinated 2 parallel workstreams — legacy support and full redesign (modern React/Angular front end plus a new CI/CD DevSecOps pipeline) — tracking dependencies, impediments, and release timelines against government stakeholder commitments.
  • Built the Jira boards and executive dashboards that gave CMS and Softrams leadership clear visibility into modernization progress, risk, and trends.
  • Founded a Scrum Community of Practice and trained leadership on Agile processes, toolsets, and AgilityHealth feedback loops.
  • Scrum Master for the front-end team designing betting terminals driven by historical horse-racing data, and for the back-end team building the pricing engine — bet pricing, customer payout calculation, and the commission charged for providing that wagering data.
  • Coordinated delivery between terminal UX and the wagering back end — where odds, payouts, and commission math are revenue-critical by definition — alongside DevOps and enterprise teams.
  • Ran dependency coordination, impediment tracking, capacity assessment, effort estimation, and delivery forecasting across 3 very different engineering groups on one cadence.
2010–2018·Earlier roles Scrum Master & Business Analyst / Capital One · Anthem · Ross · Staples · Bass Pro Shops

Waterfall-to-Agile transformations, Jira normalization, requirements gathering and JAD facilitation, UAT, release forecasting. The decade when I learned that process only matters when it ships something.

5 companies8 yearsWaterfall → Agile

03 — Independent builds

Shipped, not slideware.

Live · SaaS

Kensho Cards

NFC + QR smart business cards, in production with real users.

Founded and built solo with Claude Code: Next.js on Vercel, Neon Postgres, Clerk auth, Stripe payments. A 7-SKU line, a 1:1 QR-to-NFC encoding spec, a bilingual manufacturing RFQ package, international supplier negotiation. The whole company, one keyboard.

Next.jsVercelNeonClerkStripeClaude Code
Pre-launch

Rankora

Local SEO audits, researched and written by AI.

An AI-powered local SEO audit platform. A Python pipeline has already surfaced 148+ qualified local business leads via DataForSEO, and a custom Claude skill turns raw audit data into client-ready PDF reports with scoring and charts. Launch pending; the pipeline works.

PythonDataForSEOClaude skillsPDF reportsLead-gen
Daily driver

Multi-Agent Delivery Toolkit

9 agents and 28 skills, pointed at program delivery.

A personal AI engineering environment on Claude Code and MCP: agents automating sprint reporting, document generation, UI design, and analysis workflows. The same discipline I bring to release trains, turned on my own workload.

Claude CodeMCPAI agentsPythonAutomation

04 — Built with AI

I don't just manage AI adoption. I ship with it.

Plenty of delivery leaders talk about AI transformation; I'd rather demo it. My dashboard gives executives live delivery truth. My 9-agent toolkit drafts my sprint reports. My SaaS runs in production with real users. And this site was designed and built with Claude Code — multi-agent swarm and all — in a single session. Yes, I checked the burndown.

claude — session log — this very page
[orchestrator]+00:00session start — résumé read, swarm spawned
[content-strategist]+00:06copy shipped — zero corporate filler survived
[art-director]+00:12design system locked — contrast AA, hand-checked
[engineer]+00:31index.html written — 1 file, 0 dependencies
[reviewer]+00:38 approved — keyboard, contrast, reduced motion
[delivery]+00:41 shipped nishantpancholi.com — release readiness: green
0 external requests 1 self-contained HTML file 60fps, reduced-motion-safe

05 — Skills

The toolbox.

Delivery & Program Tooling

JiraConfluenceAzure DevOpsSmartsheetServiceNowPower BIMuralAgilityHealth

AI Engineering

Claude APIClaude CodeMulti-agent workflowsMCPPrompt engineeringPython automation

Cloud & DevSecOps

AWS GovCloudFargateEKSDockerGitLab CI/CDSonarQubeDevSecOps pipelines

Product & Engineering Context

Next.jsVercelReact / Micro-FrontendsNode.jsJava Spring BootStripePostgreSQLREST APIs

06 — Credentials

Paper trail.

Certifications

  • SAFe Release Train EngineerScaled Agile
  • Professional Scrum Master I · II · IIIScrum.org — all three levels
  • Certified ScrumMasterScrum Alliance
  • SAFe Advanced Scrum MasterScaled Agile
  • SAFe Product Owner / Product ManagerScaled Agile
  • Certified Scrum Product OwnerScrum Alliance
  • ICAgile Certified Professional — Agile CoachingICAgile (ICP-ACC)
  • Certified AgilityHealth FacilitatorAgilityHealth

Education

  • Business Management (Marketing)Lambton College — Ontario, Canada · 2008–2010
  • B.S. Computer ScienceDharmsinh Desai Institute of Technology — Nadiad, India · 2005–2008

07 — Contact

Building something that has to ship?

I'm focused on RTE, Senior TPM, and Agile program leadership roles — ideally somewhere AI-forward, where a delivery leader who builds is a feature, not a curiosity. Email is fastest; I reply like someone who tracks cycle time.

nishantpancholi@hotmail.com

Washington DC–Baltimore AreaPublic Trust

Copied.