Mohit Shrivastava logoMOHIT SHRIVASTAVAtechnology leader · builder · operatormohit@portfolio ~/career↓ download-cv
TERMINAL · mohit@portfolio ~/career
mohit@portfolio ~/career › git checkout product
Product apps

Products built from idea to store — web, native iOS and Android.

For teams that need someone who can take a product from nothing to shipped — and has done it solo, on personal money, more than once.

~20 days
idea → both app stores
Tassenger v3, from scratch
3 surfaces
native iOS · Android · web
solo — no wrappers
RM500K/yr→RM0/mo
video platform cost
build over buy
2M+
daily views served
by a product I built alone
+20%
candidate participation
Careerverse virtual fair
Let's talk →↓ download cv
why me, for this
$ cat product.md

Product work is where my founder side shows. Tassenger is the proof: a taskable messenger — chat where tasks don't die — rebuilt from scratch in ~20 days as three surfaces (native Kotlin/Compose, native SwiftUI, web console) on Supabase and Cloudflare, by directing AI coding agents under real engineering gates. It's in the stores now.

Before that: a YouTube-native video CMS that became a production system serving 2M+ daily views; immersive product launches for BMW, Nike and Netflix through Accenture Song; and a long tail of products across parking systems, dashboards, cinema platforms and NFT experiments.

What I bring to product teams is judgment about what not to build. The Video CMS exists because I refused a ~RM500K/year vendor path; Tassenger's realtime is deliberately narrow; features get cut when they don't serve the launch. Product velocity comes from saying no early and verifying everything that ships.

the business outcome

The build-vs-buy instinct, in one chart.

Product judgment is mostly knowing what not to build and what not to buy. When every enterprise video vendor quoted FMT around half a million ringgit a year, I built a YouTube-native CMS in about four weeks that now runs 33K+ videos and 2M+ daily views — for zero ringgit a month. The same instinct runs through Tassenger: deliberately narrow realtime, deferred nice-to-haves, native code where it matters. Velocity comes from saying no early, then verifying everything that ships.

Enterprise video vendor path~RM500K/yr
The CMS I built insteadRM0/mo
problems I've already solved
“We need iOS and Android, but can only staff one team.”
Real native on both — SwiftUI and Kotlin/Compose from one disciplined codebase-pair, no lowest-common-denominator wrapper.
→ Tassenger: both stores, built solo
“Every problem comes back as a vendor quote.”
Audit what the open platforms already give you, then build the thin layer you actually need on top.
→ RM0/month video platform · ~RM500K/yr avoided
“Our ideas die in planning.”
Requirements interviews → locked source of truth → mission slices → ship. v3 went from empty repo to store builds in ~20 days.
→ 21 requirement rounds · 9 ADRs · shipped
“Fast usually means fragile.”
Speed with gates: SQL test suites, UAT evidence, automated checks that reject lazy patterns before they land.
→ 58 SQL suites · 230 UAT documents · nothing ships unverified
“Customers can't feel the product before they buy.”
Immersive product experiences — 3D worlds, WebAR try-ons, virtual venues — in the browser, no headset.
→ BMW · Nike · Netflix · Careerverse (+20% participation)
“Work gets assigned in chat and disappears.”
Product insight from watching real companies: chat where any message becomes a tracked Task with an owner, state and score.
→ Tassenger — the chat where tasks don't die
first 90 days on your platform
Days 1–30
Scope ruthlessly
Lock what the product is — and the longer list of what it isn't. Requirements interviews, a source of truth, and the architecture decisions written down before serious code.
Days 31–60
Ship the spine
The end-to-end core working on every surface — ugly but real, in front of real users. Velocity with verification gates, not vibes.
Days 61–90
Harden and launch
Store compliance, review cycles, UAT evidence, the operational tooling nobody plans for — and a launch that's a release, not an event.
signals
0 → 1, repeatedly
Tassenger, Video CMS, and a 150+ project history of taking ideas to production.
Real native mobile
SwiftUI and Kotlin/Compose — not wrappers. Both stores, one person.
AI-augmented velocity
v3 in ~20 days by running AI agents like an engineering team — with verification gates.
Frugal architecture
Supabase + Cloudflare + YouTube-as-database. Spend where it matters, delete the rest.
Stores, not demos
iOS live on the App Store, Android in closed beta — releases, reviews, store compliance, the whole grind.
Trust modeled in the schema
Tassenger's org admin can't read your work — product values enforced in Postgres row-level security.
case studies on this branch
~/career › git log --branch=product
e5f6a7bYouTube-native Video CMSBuilt a production-grade, YouTube-native Video CMS that manages 33K+ videos and serves 2M+ daily views for RM0/month — replacing a ~RM500K/year enterprise-video vendor path by treating YouTube as the database.RM0/mo · ~RM500K avoided · 2M/day · 97% cache hitmediae9d2c4aImmersive, Without the HeadsetWhen lockdowns took away showrooms, career fairs and shoe stores, I led the technical delivery that recreated those experiences on an ordinary screen — virtual worlds, WebAR and digital twins for BMW, Nike, Netflix, Telekom Malaysia and Accenture itself. No headset required.BMW · Nike · Netflix · TM · +20% participationimmersiveb4c5d6eTassenger — Taskable ChatMost companies run work through WhatsApp — and the work dies in the chat. Tassenger is a taskable messenger where any message becomes a trackable Task: states, queues, proof, review, recurrence and scoring. V3 was rebuilt from scratch in ~20 days — solo, by directing AI coding agents under engineering gates.20-day v3 rebuild · iOS live · 2 native appsproduct
related projectsfull archive →
Tassenger — Taskable Chatfeatured
Own product · 2026 · Solo
Chat where tasks don't die — v3 rebuilt from scratch in ~20 days. iOS live on the App Store; Android closed beta on Google Play.
Kotlin/ComposeSwiftUISupabaseCloudflareTypeScript
YouTube-native Video CMSfeatured
FMT · 2025 · Solo
33K+ videos, 2M+ daily views, RM0/month — YouTube as the database.
Next.jsMongoDBCloudflareYouTube APIWebSub
Careerverse (virtual career fair)
Accenture Song · 2022 · Team
Virtual career fair — +20% candidate participation.
ReactBabylon.jsMicroservices
WebAR Digital Twins (footwear)
Accenture Song · 2022 · Team
AR try-on with digital twins for a global shoe brand.
8thWallWebARReact
Vehicle Parking SystemNDA
Client · — · Solo
RFID + ANPR parking automation with dynamic tiered billing.
AngularNestJSKubernetesAWSDynamoDB
DashView Pro
Own · — · Solo
Dashboard with JWT auth and dynamic data visualization.
ReactExpressPostgreSQLJWTChart.js
stack on this branch
Kotlin · Jetpack ComposeSwift · SwiftUINext.jsReactSupabase / PostgresCloudflareNode.jsTypeScriptMongoDB
● Available — serving notice · open to interviews now

Building a product and need it shipped?

If you need a product taken from idea to shipped — I've done it solo, on my own money, more than once. Imagine what I'd do with your team.

Let's talk on WhatsApp →↓ Download CV