Portfolio · Product / Delivery

I make the call, then I write down why.

Dipika Chawla — product leader working inside regulated, operationally complex systems: Medicaid, Medicare, and CMS-compliant platforms. The case studies below are decisions, not deliveries — what I chose, what I turned down, and what it cost.

Focus: Regulated healthcare platforms Background: Engineering → Product Currently: Applied AI / workflow tooling
Case Studies
Three decisions, with the option I didn't take

Each entry follows the same log: the constraint, the paths on the table, the one I picked, and what it cost to pick it.

CASE — BILLING-01 Shipped
Interim automation over waiting for the "right" fix
CMS-Compliant Medicaid Waiver Platform · Product Manager
Constraint
Referral claims processing was consuming a disproportionate share of manual operational time inside the CMS-compliant Medicaid waiver platform — well ahead of when full system integration was scheduled to land.
Options on the table
  • Wait for full system integration
  • Add claims-processing headcount to absorb the load
  • Scope a narrow, interim automation layer for referral claims only
Cost of the choice
Traded a fully generalized long-term solution for speed. Accepted that the interim layer would need to be retired once full integration shipped — deliberate technical debt, scoped narrowly enough not to entangle with the future build.
95%
Manual workload reduced
CASE — INTAKE-02 Shipped
Direct integration over pushing standardization onto partners
CMS-Compliant Medicaid Waiver Platform · Product Manager
Constraint
Referral intake ran through fragmented external Social Care Network portals and internal systems, with no shared format and no standard entry point.
Options on the table
  • Standardize manual intake across all partner portals
  • Ask each SCN partner to adopt a common format first
  • Build direct API integrations against each system as it existed
Cost of the choice
Gave up the simplicity of one shared standard. Took on integration and maintenance work across multiple fragmented external systems rather than waiting on partners who had no real incentive to standardize first.
85%
Manual processing time reduced
CASE — STORYPILOT-03 In build
A conversion layer, not a replacement
StoryPilot · Founder / Sole PM
Constraint
Turning a BRD into user stories, Jira tickets, and Confluence docs is manual and repetitive, and context gets rebuilt from scratch at every handoff between formats and people.
Options on the table
  • A single-shot BRD-to-ticket generator with no memory
  • A full project-management tool replacing Jira and Confluence
  • A persistent-context layer that plugs into existing tools
Cost of the choice
Gave up the bigger story of "replace your PM stack" for a smaller, faster-to-ship one — and took on the harder adoption problem of proving it's worth using alongside existing tools, not the easier problem of being the only tool.
React
+ applied prompt architecture
Experience
Where the delivery track record lives

The rest of the resume — timelines hit, teams run, systems shipped.

Nov 2024 — Present
Product Manager
Healthcare Engagement Platform — Medicare / Medicaid workflow, e-commerce, and messaging
Sept 2022 — Nov 2024
Senior Software Engineer
Candidate Engagement Platform — full-stack ownership, AWS
2004 — 2022
Founder & CEO
Cushyfoods LLC — one to three locations, full P&L ownership
1999 — 2003
Technical Lead
Sprint E|Solutions, BroadVision, Black & Veatch
Capabilities
What sits underneath the decisions
Product
Roadmap development, backlog prioritization, acceptance criteria, business case development
Healthcare IT
FHIR, HL7, Mirth Connect, CMS compliance, Medicaid waiver programs
Analytics
SQL, PostgreSQL, Power BI, A/B testing, KPI definition
Build
Python, REST APIs, AWS, Azure, React, C#/.NET, TypeScript