Back to portfolio
Production workGrowth & monetizationAnalytics & automation

Operating evidence from products I shipped.

I joined AlloHealth's Founder's Office to work close to the P&L, then took ownership of Repeat Business: the full customer journey after a first purchase. The common thread is measurable customer and business outcomes, not feature output.

33% → 45%

Repeat revenue contribution

₹6.5K → ₹7.5K

Realized LTV

35% → 85%

Doctor-agent handoff

21.7% → 16.7%

Payout cost / revenue

01 · Repeat business

Made the post-purchase journey one owned system.

The business needed repeat revenue to become a larger, healthier share of total revenue. The largest leakage happened before the first follow-up, while ownership was split across fulfilment, treatment support, doctors, WhatsApp, and calling teams.

01First purchase
02Medicine & test fulfilment
03Treatment support
04Follow-up intent
05Booking & attendance
06Repeat fulfilment

My ownership

Led a seven-person pod and aligned product, operations, doctors, fulfilment, and customer experience around one measurable repeat funnel.

What changed

Improved checkout and fulfilment, treatment communications, follow-up booking channel drop-offs, attendance, reschedules, no-shows, and the loop into repeat fulfilment.

How it was managed

Built a City Heads dashboard that translated funnel gaps into the incremental repeat consultations each doctor and city could unlock.

Business resultRepeat revenue contribution increased from 33% to 45%; realized LTV increased from ₹6,500 to ₹7,500 in the latest-quarter view versus the prior two-quarter baseline.

02 · Conversion

Closed the gap between consultation and purchase.

After an online consultation, patients needed a reliable handoff from the doctor to an agent who could explain the prescription and support purchase. Doctors leaving early or inconsistent handoffs created avoidable conversion loss.

Before35% handoff

Inconsistent transfer after the first consultation.

After85% handoff

Pop-up, timer, and workflow guardrails kept ownership clear.

  • Designed changes in the doctor product and the agent workflow, not just another operational reminder.
  • Applied across roughly 5,000 eligible patients per month.
  • Patients receiving the handoff converted to purchase at about twice the usual rate.

03 · Payout platform

Turned doctor compensation into a product.

The goal was to reduce doctor payout cost as a percentage of revenue without breaking incentives or creating an opaque finance process. I designed a declining, slab-based model and the system required to calculate and operate it.

1Revenue inputs
2Tiered rules
3Payout calculation
4Approval
5Statement & payslip
6Payment trigger
4 weeks

Idea to production

With 1 engineer, 1 designer, and 1 analyst intern

60+

Doctors covered

Automated statements, payslips, and approvals

₹25L / month

Transaction volume

Final controlled action initiated payments

Economic resultDoctor payout cost fell from 21.7% to 16.7% of revenue while the product made calculation, approval, and payment operations more consistent.

04 · AI & operations

Used AI to inspect the funnel, then kept humans in control.

The WhatsApp chatbot handled about 7,000 chats a month. I used n8n- and Langfuse-based workflows to audit conversations, classify follow-up journeys, surface incorrect resolutions, and recommend the next product or content fix. The team reviewed the evidence and made the decisions.

Conversation data
AI-assisted audit
Human review
Journey fix & measure
Daily data-to-decision pipelineScheduled queries → source validation → safe Sheet repair → dashboard rebuild → guarded publication

The refresh checks both the Google Sheets sources and published City Heads dashboards for staleness. If a query-backed range is incomplete, it validates the latest exact-name scheduled-query output, writes it to a staging tab, verifies it, and replaces only the owned columns. Helper formulas and unrelated tabs stay untouched; a stale or structurally invalid source blocks the dashboard publish.

35% → 55%

Patient-led bookings

Higher self-serve share across total follow-up bookings

88%

High-intent self-booking

Only 12% of intent cases needed agent help

9K → 6.5K

Weekly calling tasks

Deduped tasks while carrying forward customer context

05 · P&L & fundraising

Worked where product decisions meet the books.

Before leading Repeat Business, I shadowed the company P&L owner and supported the pre-Series A process led by Rainmatter. This built fluency in how operating metrics become financial reporting and investor diligence.

Investor-grade analysis

Prepared one-year cohorts, LTV, indirect P&L, and supporting data after the term sheet in close partnership with the Chief of Staff.

Books-aligned reporting

Worked directly with CAs and consultants on the financial numbers used for annual reporting and investor communication.

Gross-margin improvement

Improved margin by about 27% across three high-velocity SKUs through stockout control, vendor changes, and better pricing.

Why this transfers to fintech

Different domain. Familiar product problems.

I am targeting B2C product roles in lending, payments, trading, and wealth where analytics, trust, and operating economics matter.

Evidence I haveFintech applicationShared product skill
Repeat treatment journeyRepayment, renewal, and lifecycle journeys

Both require timely nudges, clear customer state, exception handling, and trust.

Doctor payout platformSettlements, commissions, and partner payouts

Tiered rules, approvals, statements, reconciliation, and payment initiation share the same core system needs.

P&L and cohort ownershipUnit economics and credit portfolio thinking

Cohorts, LTV, margin, cost-to-revenue, and investor-grade reporting anchor product trade-offs in economics.

AI-assisted conversation auditsServicing, collections, and support quality

High-volume conversations need traceable review, human judgment, prioritization, and measured iteration.

Independent markets experience

Two years of real capital, not only classroom interest.

During college, I managed a ₹10L pledged-capital trading account to approximately ₹16L over two years. I independently studied equities, derivatives, bonds, REITs, market structure, and risk. This is presented as personal markets experience—not employment, a managed fund, or investment advice.

More proof

Resume, data work, code, and a complete fintech concept.

All company metrics are presented at an aggregate level. No patient, doctor, or confidential company data is included.