Skip to content
MP Thaker

I build systemsthat turn expertise and AIinto operating capability.

I am a founder, operator, and product leader. I build working systems, stay close to the people using them, and change the plan when the evidence disagrees.

Building now

Grapevines

I built Grapevines solo, end to end. It combines adaptive conversation, research, voice calibration, and structured evaluation so a professional can reason through a career move before generating materials.

Official UI · fictional recordsCaptured from Grapevines’ current public demo interface with controlled fictional record substitutions.

The working product joins adaptive conversation, research, voice calibration, and a scoring pipeline in one operating system.

Selected builds

See how the systems make decisions.

Little FablesPublic demo

An approved story becomes an illustrated, narrated reader for day and bedtime.

Little Fables turns an approved custom story into an illustrated, narrated private reader with day and bedtime modes. There are no runtime model calls or child-facing accounts.

Read the case
Public demoCaptured from Little Fables’ intentionally public demo. No buyer intake, child photo, household record, or private custom-order source material appears here.
HummingbirdPrivate workbench

A financial operations system that keeps execution behind a human decision.

Hummingbird connects financial data, analysis, and paper-trade execution inside a private workbench. Its agents observe, propose, and execute in separate roles so automation never quietly becomes authority.

Read the case
Synthetic recordsCurrent Hummingbird interface captured from an isolated copy with fixed synthetic API responses. No real account connection or owner record was used.
CanairyPreparedness prototype

Public risk signals translated into decisions a household can actually make.

Canairy watches public economic, supply, energy, and geopolitical signals, then translates them into calm household actions. The product is about interpretation and timing, not a louder stream of alerts.

Read the case
Simulated scenarioA deterministic sample scenario, not live conditions and not a forecast.

Operating record

The work behind the systems.

Coqui Coqui

Director of Customer Experience

Hospitality operations across the customer path, from brand and booking through arrival and service.

Subziwalla

Co-Founder and Head of Product

Built and operated a grocery marketplace from its first version through the company’s exit.

Panso

Co-Founder and Head of Product

Led product work for a customer-experience platform spanning transaction data, campaigns, and enterprise delivery.

Lovingly

AI Product Manager

Worked on marketplace re-platforming and applied AI workflows with explicit evaluation and human review.

Read the case

Writing

Notes from inside the work.

See all writing

MP Thaker

About

I learned systems by owning the handoff.

My operating path runs through hospitality, a grocery marketplace, a customer-experience platform, marketplace re-platforming, and products built with AI. The industry changed. The responsibility did not: make sure the next person has the context to act.

I am most useful where a capable team has access to powerful tools but still needs someone to decide what should be built, what evidence counts, and who owns the result.

Contact

Start with the operating question.

If you are building an AI-enabled capability and the hard part is ownership, judgment, or execution, tell me where it is stuck.