Harshith
Beecha.

Hey, I'm Harshith

I started by measuring a product. Now I build the platform I used to measure.

Behind every user story,
there's an actual person.
I start with who. Then I ask why.

Associate PM at Conversive (SMS-Magic). I started as a Product Analyst. Instrumented the product, found where users struggled, fixed it. Each fix led to bigger ownership: onboarding, the omnichannel platform, voice AI. Scroll for the story.

Read the story ↓View ResumeAbout me
3.5
years experience
3+
products shipped
30+
enterprise customers
75%
onboarding time cut
01

The Work

Growth · Onboarding Optimization

Cutting Salesforce onboarding from 60 to 15 minutes

Sixty minutes was way too long.

75%
Onboarding time cut
OnboardingData-drivenSalesforceMixpanel
B2B SaaS · Omnichannel Platform

Building the conversation object that became Conversive's platform foundation

0→30 enterprise customers in 7 months, built on the data model I spec'd.

0→30
Enterprise customers
0→1 BuildOmnichannelB2B SaaSAI Routing
AI · Enterprise Voice Platform

Cutting enterprise agent load by 40% with AI voice

Turns out, AI can handle that call.

40%
Agent load reduced
AI / VoiceEnterpriseIVRHealthcare
02

Building on the Side

Looply

Live

Turns customer feedback into prioritized product insights.

Synapse

Building

AI-powered second brain for developers. Paste code, ask questions, get answers.

URLzippy

Live

Smart link shortener with analytics and UTM auto-tagging built in.

03

Products I Think About

Pricing Strategy

Intercom: Fin AI Pricing

How Intercom made customers pay per resolved conversation, and why that changes everything.

Outcome-based pricing for Fin AI. Charges per resolved conversation, not per seat. Escalation rate as the north star metric. What this means for the PM who owns pricing.

Read teardown ↗
Product Critique

Linear: The "Why" Layer

Linear is the best project tool I've used. It's also missing something important.

Opinionated design as a growth strategy. What Linear does brilliantly, and the one layer that's still missing: no native goal or outcome tracking. A PM perspective on the gap.

Read teardown ↗
04

The Person Behind the PM

I think in problems, not features.

I joined SMS-Magic as a Product Analyst. My job was to make the product measurable, and the measurements kept pointing at things worth fixing. Three years later I own platform workstreams: onboarding, omnichannel, voice AI.

I build in Hyderabad, usually with small teams, short cycles, and real customers on speed dial.

Fall in love with the problem. Don't get infatuated with your own solutions.

Uri Levine, co-founder of Waze
Get in touch →
🎯

Customer-close by default

I've spent as much time in the support ticket backlog and on CS/O&I calls as in the PRD. I don't ship things I haven't watched a customer struggle with first.

📊

Data first, story second

I look at the numbers before I form an opinion. Then I go find the human reason behind them.

I'd rather ship and learn

Perfect specs don't ship. I'd rather try something, break it in front of real users, and fix it fast.

05

How I Got Here

Chapter 1

The instruments

I joined SMS-Magic as a Product Analyst with one mandate: make the product measurable. I built our Mixpanel infrastructure from scratch: 16+ dashboards, tracking standards, and user journey analysis that showed exactly where the experience was breaking. Fixing those friction points was my first product work; NPS moved because of it.

One metric I'd instrumented, time-to-first-message, found the biggest problem.

Chapter 2

The 60-minute wall

The data showed new customers taking 60+ minutes to send their first message. I redesigned the entire Salesforce onboarding flow: 60 minutes became 15.

60→15 minRead the full story

That work earned me the APM seat. To decide what to build next, I went where the unfiltered signal was: the support ticket backlog. Analyzing it for quarterly PI planning kept surfacing the same pattern: channels that didn't talk to each other.

Chapter 3

Rebuilding the core

Conversive's channels were islands. I led the 0→1 build of the conversation as a first-class object, the data model the whole platform now runs on. 0→30 enterprise customers in 7 months.

0→30 customersRead the full story

Then the company acquired a voice AI product, and the conversation object I'd spec'd had to absorb an entire new synchronous channel.

Chapter 4

Voice on the foundation

Voice AI went from out-of-scope to flagship. I defined the use cases, the seven quality requirements, and the eval framework that made it enterprise-grade. 40% reduction in agent handling time across 10+ accounts.

40% agent load cutRead the full story

Analytics taught me to find the problem. Tickets taught me to hear it in the customer's words. Product taught me to fix it at the root. That's the loop I run on everything.