Skip to content
Canada — RemoteAI Products · Full-Stack · Data

Mikky

Product Engineer | Product Leader

I build analytics platforms — and I own the whole loop: talk to users, decide what's worth building, write the code, ship it, instrument it, and read the data afterward.

View the workSee the data lab

//About

I build the layer between raw data and the decision.

I own the whole loop: talk to users, decide what's worth building, write the code, ship it, instrument it, and read the data afterward.

I'm a product engineer with 15+ years across backend engineering, cloud infrastructure and product leadership. What's stayed constant is the shape of the work: I sit on the discovery call, decide what's actually worth building, write the code, ship it, instrument it, and then read the data to find out whether I was right.

That loop has produced real numbers. I prototyped a product concept, built the POC, and shipped the MVP that grew into a line generating $25M+ a year — which earned Innovator of the Year in 2023. Across launches I've contributed $45M+ in new product-derived revenue and 40% ad-revenue growth, and I've taken 10+ products from 0 → 1 for Fortune 500 and Tier-1 customers.

My centre of gravity is data. I write the SQL and build the instrumentation behind product KPIs rather than filing a ticket with a data team, and I've built analytics and reporting layers from scratch — Superset dashboards, SQL models, Grafana and Datadog telemetry — that replaced manual reporting and now drive roadmap decisions. I care obsessively about correctness: a chart that's beautiful and subtly wrong is worse than no chart at all.

Most recently that's meant AI: an agentic product built on GenAI, predictive modelling and MCP-based tooling, plus AI workflows that measurably cut delivery cycle time across engineering and product.

Also — my full name is Bhaumik Shukla, which fifteen years of meeting intros have taught me is a genuine pronunciation hazard. People usually get two syllables in, improvise the rest, and arrive somewhere entirely new. So I saved everyone the suspense: Mikky.

Own the whole loop

Discovery, definition, build, ship, instrument, then read the data. Handing off half-understood requirements is where products die.

Correctness is a feature

Every number is traced from source query to rendered pixel before it ships. A wrong chart costs trust you don't get back.

Instrument, don't guess

+30% engagement and −40% time-to-first-value came from instrumentation-led iteration, not opinions in a planning meeting.

Prototype before you commit

I validate concepts in days with AI-assisted development before spending real engineering cycles on them.

15+

Years shipping products

Backend, cloud infra and product leadership

$45M+

New product-derived revenue

Across launches, plus 40% ad-revenue growth

10+

Products taken 0 → 1

Discovery to production, Fortune 500 & Tier-1

3×

Awarded for innovation

Three innovation and revenue awards

Figures above are drawn from my résumé. Anything marked provisional elsewhere on this page is an estimate pending confirmation.

//Skills

Fifteen years of range, plotted honestly.

Six competency axes rather than a wall of logos. The bars are the same numbers as the radar — read whichever you prefer.

Competency profile, self-assessed

Six axes, 0–100. Deepest in data and backend; frontend is deliberate rather than decorative.

Data & Analytics95
Backend & Languages93
AI & Agentic88
Cloud & SRE90
Product92
Frontend82

Data & Analytics

  • SQL modeling
  • Apache Superset
  • Metabase
  • Amplitude
  • Elasticsearch (ELK)
  • Looker
  • Tableau
  • Power BI
  • ETL pipelines
  • Kafka
  • Spark
  • Snowflake

Backend & Languages

  • Python
  • TypeScript
  • Node.js
  • Go
  • Java
  • Scala
  • SQL
  • Django

AI & Agentic

  • LLM APIs (Claude, OpenAI, Gemini)
  • Agentic systems
  • MCP
  • RAG
  • Prompt & eval design
  • Predictive modeling
  • Vision AI

Cloud & SRE

  • GCP
  • AWS
  • Azure
  • Kubernetes
  • Docker
  • Terraform
  • CI/CD
  • Grafana
  • Datadog
  • Incident response

Product

  • Customer discovery
  • Product definition
  • KPI instrumentation
  • Roadmapping
  • Rapid prototyping (Figma)
  • Agile / Scrum

Frontend

  • React
  • TypeScript
  • Next.js
  • Vue.js
  • Charting / D3
  • HTML / CSS

//Data Lab

Rather than claim I know data, here is the work.

Every chart below is built the way I build them in production: one axis, a validated colorblind-safe palette, a legend whenever there are two series, and a data table behind every figure so nothing is ever encoded in color alone.

Colors pass CVD separation and contrast checks — computed, not eyeballed

Ordinal stages get one hue in ordered steps; nominal categories share a hue

Every figure has a “View data” table — the accessibility relief channel

Deterministic sample data, so no hydration mismatch on a static build

Illustrative data, generated deterministically — shaped like the real thing, not real customer data.

Daily active accounts

3,412

+12.4%vs. prior period

Queries / minute

1,876

+8.1%vs. prior period

p95 query latency

612 ms

−18.6%vs. prior period

Pipeline freshness

42 s

−9.2%vs. prior period

Dashboard usage, trailing 12 months

Sessions and analytical queries both compounding — queries growing faster, which is the signal that people are actually exploring rather than glancing.

  • Sessions
  • Queries run

p95 query latency by dashboard surface (ms)

Sorted heaviest-last. Raw event search is the outlier that justified a pre-aggregation layer.

Dashboard activity by day and hour

A business-hours ridge on weekdays — the usage shape of a tool people work in, not one they check.

Bid pipeline, request → impression

Where supply actually falls out. Each stage is a place where an SSP integration either holds or leaks.

//Selected work

Two platforms I took from a customer conversation to production.

Most recent first. Each one is the whole arc — discovery, definition, build, delivery, and the feedback loop afterward.

Enterprise AnalyticsAI Image Recognition & Broadcast Media Advertising Platform · Jul 2025 — Present

In-house Analytics & Reporting Platform

Built the analytics layer from scratch — React and TypeScript on GCP reading production MySQL — and turned it into the dashboard Fortune 500 and Tier-1 teams open every morning.

4×

Adoption growth

Expanding into media and OTT

22%

Of company revenue

From 5+ Tier-1 clients won

100%

Manual reporting replaced

Superset + SQL models took over entirely

<2s

Median dashboard load

On multi-million-row datasets

Provisional

I built the company's analytics product in-house rather than buying one: a React and TypeScript application on GCP that reads directly from MySQL and renders enterprise-grade charts, drill-downs and reporting for large customers.

The hard part was never drawing the chart. It was getting the data model right so the same number means the same thing on every screen, at every level of aggregation, for every customer's slice of the data. I designed the SQL models, the transformation layer and the rendering layer as one system, so correctness holds end to end — then wired Grafana and Datadog instrumentation underneath it.

It replaced manual reporting outright, and it now drives roadmap decisions rather than just describing the past.

This is a startup, so the role came with every hat: I ran the customer conversations, wrote the product definition, laid out the development plan, built the product myself, worked with the team on the connective pieces, delivered it, and then owned the feedback loop — shipping improvements as real enterprise usage exposed friction.

Alongside the in-house build I delivered embeddable analytics through Apache Superset and Metabase, so customers who wanted our dashboards inside their own products could have them without us rebuilding the surface each time.

What I owned

  • Customer discovery and ongoing product-level communication with Tier-1 accounts
  • Product definition and development planning from a vague enterprise ask
  • Full-stack build: React + TypeScript frontend, data layer, GCP deployment
  • SQL modeling and query design against production MySQL
  • Data-level verification at every stage — query, transform, aggregate, render
  • Embeddable dashboards via Apache Superset and Metabase
  • Instrumentation and observability with Grafana and Datadog
  • Post-launch feedback loop and continuous improvement

Stack

  • React
  • TypeScript
  • Python
  • SQL
  • MySQL
  • GCP
  • Apache Superset
  • Metabase
  • Grafana
  • Datadog

Adoption trajectory

Illustrative

Shape of the 4× adoption growth as the platform expanded into media and OTT.

Screenshot slot

public/projects/enterprise-analytics-dashboard/drilldown.png
Drill-down view
Performance Overview dashboard: KPI tiles for total interactions, engaged sessions, engagement rate, unique users and conversion rate, above a geographic activity heatmap, a combined interactions-and-engagement time series, a device-breakdown donut, per-segment sparkline cards and a content-collections table.
Performance Overview — KPI row, geo activity heatmap, engagement trend, device split and per-segment breakdownsView full size ↗
Ad Tech PlatformAI Vertical Video Platform · Jul 2025 — Dec 2025

SSP Integration & Onboarding Platform

The advertising backend for a vertical video platform — enterprise onboarding plus every supply-side integration, delivering ads across multiple channels.

5+

Tier-1 clients onboarded

Enterprise partners live in production

12+

SSPs integrated

Supply-side partners in production

Provisional

4×

Faster onboarding

Customer go-live vs. the manual process

Provisional

99.9%

Delivery uptime

Ad serving availability

I built the advertising backend for a vertical video platform: the system where enterprise customers get onboarded, every relevant supply-side platform gets integrated, and ads then deliver across multiple channels.

It was the most complex surface I've worked on. Each SSP arrives with its own contract, its own quirks and its own failure modes — and the platform has to normalise all of that into one predictable onboarding and delivery path that a Tier-1 customer can trust in production.

I owned it at the product level as well as the code level: running the customer conversations, translating what enterprise partners needed into a concrete integration plan, building it, and carrying it through to delivery and production support.

What I owned

  • Product-level customer communication with enterprise and Tier-1 partners
  • Architecture of the SSP integration layer and onboarding flow
  • Implementation of multi-SSP integrations and multi-channel ad delivery
  • Normalising heterogeneous SSP contracts into one predictable pipeline
  • End-to-end delivery ownership and production support

Stack

  • Python
  • TypeScript
  • Node.js
  • OpenRTB
  • VAST
  • SQL
  • GCP

Bid pipeline, request → impression

Where supply actually falls out. Each stage is a place where an SSP integration either holds or leaks.

Screenshot slot

public/projects/ssp-integration-platform/console.png
Onboarding console — drop your screenshot here

More case studies in progress

//Experience

Fifteen years, six companies, one consistent job description.

Backend engineering, cloud infrastructure and product leadership — usually all three at once.

  1. AI Image Recognition & Broadcast Media Advertising Platform

    CurrentJul 2025 — Present · 1 yr 2 mos

    Senior Product Engineer — AI, AdTech & Data Platform

    Canada — Remote

    AI-based image recognition and broadcast media advertising platform.

    • End-to-end ownership of customer-facing AI and data products — discovery through production.
    • Ship customer-facing AI and data products end to end — discovery, scoping, implementation with engineering, and production instrumentation.
    • AI and agentic workflow design and implementation.
    • Built the analytics and reporting layer from scratch: Superset dashboards, SQL models, and Grafana/Datadog instrumentation that replaced manual reporting and now drives roadmap decisions.
    • Analytics and instrumentation layer across SQL, TypeScript, GCP, Superset, Grafana and Datadog.
    • Designed and implemented AI and agentic workflows that measurably improved delivery speed and decision quality across the team.
    • Rapid prototyping and technical validation ahead of build — AI-assisted workflows and AI-powered rapid prototyping.
    • Advertising / ad-tech platform creation and prototyping.
    • Drove 4× adoption growth expanding into media and OTT; won 5+ Tier-1 clients contributing 22% of company revenue.
    • Python
    • TypeScript
    • SQL
    • GCP
    • Superset
    • Grafana
    • Datadog
    • LLM APIs

  2. AI Vertical Video Platform

    Jul 2025 — Dec 2025 · 6 mos

    Product Engineer — Advertising Platform & SSP Integrations

    Canada — Remote

    AI-based vertical video platform.

    • Built the advertising backend for a vertical video platform: enterprise customer onboarding, supply-side platform integrations and multi-channel ad delivery.
    • Owned product-level communication with enterprise and Tier-1 partners, translating their requirements into a concrete integration plan.
    • Normalised heterogeneous SSP contracts — each with its own quirks and failure modes — into one predictable onboarding and delivery pipeline.
    • Carried the platform through to delivery and production support.
    • Python
    • TypeScript
    • Node.js
    • OpenRTB
    • VAST
    • SQL
    • GCP

  3. Real-Time Advertising Platform

    Jun 2018 — Jun 2025

    Product Engineering Lead (VP, Products & Technology)

    Canada — Remote

    • Owned product and engineering for a real-time bidding platform where p99 latency was a hard product constraint — shipped work that lifted engagement 30%, cut setup time 40% and grew revenue 40%.
    • Designed and built an agentic AI product for internal and customer workflows using GenAI, predictive modeling and MCP-based tooling.
    • Drove architecture across 10+ products — data models, integration patterns, build-vs-buy, and where to spend the latency budget.
    • Wrote the SQL and built the instrumentation behind product KPIs rather than requesting dashboards from a data team.
    • Implemented platform-wide privacy and compliance architecture: GDPR, CCPA, TCF 2.0, IAB political-ads protocols and consent management.
    • Python
    • Java
    • Kafka
    • Spark
    • PostgreSQL
    • Aerospike
    • Kubernetes
    • AWS

  4. Sales Intelligence SaaS

    Mar 2018 — Dec 2018

    Product & Infrastructure Engineering Consultant

    Singapore

    • Architected CI/CD pipelines and cloud deployments, cutting release time 60%.
    • Built cloud-native reliability and monitoring tooling; raised uptime to 99.9%.
    • Partnered with the CTO to diagnose platform bottlenecks, improving product efficiency 35%.
    • Docker
    • Kubernetes
    • AWS
    • Python
    • Scala
    • React

  5. Product Engineering Services

    Sep 2016 — Mar 2018

    Solution Engineering Manager

    India

    • Directed delivery of SaaS and on-prem log analytics and monitoring products for large-scale distributed systems, serving Fortune 500 customers.
    • Built operational dashboards on Elasticsearch (ELK) for infrastructure telemetry and product decisions.
    • Elasticsearch
    • Kubernetes
    • Docker

  6. Connected Health Startup

    Apr 2011 — Sep 2016

    Product Lead, Cloud Products ← Sr. Software Engineer

    Connected Health & Cloud Platform

    • Led product and architecture for cloud-based connected-health products across mobile and backend — microservices design, scalability planning and cloud strategy.
    • Owned compliance implementation for HIPAA, FDA and CE (EEA).
    • Designed and shipped REST APIs, data pipelines and analytics dashboards surfacing product usage and system performance.
    • Filed technology patents in connected health.
    • Python
    • Django
    • Node.js
    • MongoDB
    • Redis
    • AWS

Education

  • M.S., Information Technology

    2009 — 2011

  • B.C.A., Computer Science

    2006 — 2009

Awards

  • Innovator of the Year

    Ad-tech platform, 2023

  • Revenue Driver Award

    Ad-tech platform, 2022

  • Champion Award

    Connected health, 2013

Certifications

  • Digital Product Management

    UVA Darden

  • Certified Product Manager

    Product School

  • Data-Driven Planning

    The Trade Desk

//Contact

If you're building something where data has to be right, let's talk.

Product engineering, analytics platforms, AI products, or a hard integration nobody wants to own. I read everything.

mickeyconnects@gmail.com
whoami
{
  "handle":     "Mikky",
  "role":       "Product Engineer | Product Leader",
  "focus":      "AI Products · Full-Stack · Data",
  "based":      "Canada — Remote",
  "years":      15,
  "owns":       ["discovery", "definition",
                 "build", "delivery",
                 "instrumentation"],
  "deepest_in": ["Python", "SQL", "TypeScript",
                 "analytics", "cloud"],
  "open_to":    "interesting problems"
}