Product Manager · AI · EdTech · Fintech
Building products at the intersection of human experience & technology
AI-native Product Manager and Product Engineer with a passion for building and refining products that sit at the intersection of education and fintech. I lead with a deep focus on user experience and company growth, bridging technical teams and business strategy to create products that matter.
About & Background
I'm a Product Manager at the intersection of AI, fintech, and EdTech, with a background that most PMs don't have. I've been a software engineer, a director of technology, a director of training and education, and a university lecturer on three continents. All of that feeds directly into how I build products.
At Tradeify, I led end-to-end product development for TradeSage, an AI-native trading analytics and performance platform. I owned everything from initial vision and architecture through UX strategy, engineering collaboration, broker integrations, and go-to-market planning.
Before that, at Apex Trader Funding, I held several roles across the company. As the PM who led LMS selection and scaled the company's learning platform, I served the full workforce across multiple departments. As Director of Technology, I oversaw infrastructure and tooling for a rapidly scaling fintech operation. And as Director of Training and Education, I designed curriculum and training programs across departments that measurably improved operational efficiency and workforce competency.
I think AI-native product management is genuinely different from traditional PM work, and not just because the tools have changed. The role itself is shifting. When AI can generate solutions faster than most teams can evaluate them, the most valuable thing a PM brings is not speed or output. It is judgment. Knowing which problem is worth solving, which signal to trust, and where the model is confidently wrong. The PMs who treat AI as a layer on top of existing processes will be outpaced by the ones who rethink the process entirely around what AI actually changes. That is the work I find most interesting, and where I believe the field is heading.
My graduate work at Columbia University in Applied Linguistics gave me a deep understanding of how people construct meaning, frame problems, and communicate across different contexts. Language and code share structural logic, and that background made picking up programming more intuitive. Both inform how I work across technical and human sides of a product.
I completed a Graduate Certificate in Front-End Web Development at Harvard Extension School, and have hands-on experience in Python, JavaScript, SQL, Django, and Edge ML training. I completed an AI and full-stack development bootcamp at Le Wagon in Berlin, Germany and an AI Product Management certification at General Assembly, both of which sharpened how I think about building and shipping AI products.
The throughline across all of it is range. Teaching, building, leading teams, studying language, working across cultures. Each discipline has informed the others in ways that are hard to separate. That breadth is what I bring to product work.
Beyond the Work
Before technology became my medium, people were. I spent years teaching across the world: ESL in Tokyo, the West Bank, and Germany; linguistics and pedagogy at the university level in Turkey. Living and working inside those cultures shaped the way I listen, communicate, and build. I've traveled to over 30 countries, studied five languages, and that experience still shows up in how I think about users and cross-functional teams.
Outside of work I stay grounded through movement. Certified yoga teacher, dancer, flow arts practitioner, rollerskater, avid outdoors person. Discipline in one area tends to carry into everything else.
Focus Areas
AI/ML Products
Fintech Platforms
EdTech & Learning Systems
SaaS & B2B Products
Education
Columbia University, MA
Harvard Extension School
General Assembly, AI PM
Le Wagon, AI Dev Bootcamp
Portland State University, BA
Skills
Jira · Figma · Miro · Notion
Claude · ChatGPT · Cursor · OpenAI
SQL · Python · JavaScript
Agile · Scrum · OKRs
Wireframing · Prototyping
Location
Remote
Selected Work
Fintech · AI · 0 to 1
Led end-to-end product development for an AI-powered trading analytics and performance platform, building from vision to go-to-market at Tradeify.
Read Case Study →EdTech · Platform · Analytics
Scaled a company-wide learning platform serving the full workforce across multiple departments, measurably improving knowledge outcomes through data-driven iteration.
Read Case Study →Consumer · SaaS · AI · Social
A private, invitation-based social recipe platform letting families and communities create shared living cookbooks, with an AI layer and premium physical book upsell.
Read Case Study →E-Commerce · Web Dev · Publishing
Designed and built a fully functioning e-commerce and author platform for an independent publisher, including a live store, photo journal, waitlist system, and email capture.
Read Case Study →AI · EdTech · B2B SaaS · Venture Concept
A venture concept for an AI agent platform that transforms how organizations onboard, upskill, and retain talent — built from domain expertise in L&D and fintech.
Read Case Study →Work Samples
Browse all documents and deliverables directly. Filter by type or search by keyword.
TradeSage
Product requirements for the core MVP — user stories, personas, success metrics, and acceptance criteria.
View Document →TradeSage
Business case for the relaunch — objectives, market opportunity, stakeholder requirements, and constraints.
View Document →TradeSage
Phased build plan from discovery through differentiator features, with success metrics per phase.
View Document →TradeSage
Full specification for the AI coaching engine — capabilities, technical approach, edge cases, and success metrics.
View Document →TradeSage
Sprint plan with goal, backlog, planning breakdown, ceremonies, and definition of done.
View Document →TradeSage
Priority framework mapping 50+ features across must-have, should-have, and differentiator tiers.
View Document →Apex LMS
Four-phase rollout plan for an enterprise learning platform — from discovery through optimization.
View Document →Apex LMS
Three objectives with key results tied to platform adoption, knowledge quality, and business alignment.
View Document →Apex LMS
Sprint example with goal, backlog, planning session breakdown, and definition of done for a learning track build.
View Document →Griffin's Kitchen
Product requirements for a subscription recipe and meal-planning platform — personas, user stories, and metrics.
View Document →Griffin's Kitchen
Phased go-to-market roadmap from MVP launch through community and monetization expansion.
View Document →Griffin's Kitchen
Detailed spec for the platform's core features — requirements, edge cases, and acceptance criteria.
View Document →Griffin's Kitchen
Objectives and key results for the launch quarter — growth, engagement, and retention targets.
View Document →Get In Touch
I'm actively exploring Senior PM and Staff PM roles in AI, EdTech, and Fintech. If you're building something ambitious and need a product leader who bridges technical depth with human-centered thinking, I'd love to connect.
I respond to every message. The best way to reach me is via email or LinkedIn.
Remote
This is an AI representation of Lyndsey Reed. 10 messages remaining today.
Experience · Education · Skills
Summary
AI-native Product Manager and Product Engineer with a passion for building and refining products that sit at the intersection of education and fintech. I lead with a deep focus on user experience, company growth, accountability and empathy, owning end-to-end product development from vision and roadmap through execution and go-to-market. With hands-on engineering experience, I bridge technical teams and business strategy to build products that deliver value to users and create impact for the business.
Education
Experience
Technical Skills
Fintech · AI/ML · 0 to 1 Product
The Problem
Most traders on funded platforms could see their raw numbers but couldn't extract meaning from them. They had no way to identify behavioral patterns driving losses, no psychological tracking, and no personalized coaching. The tools that existed were either too generic or too complex. The gap was clear: an AI-native performance layer built specifically for how traders actually think and fail.
Product Strategy
TradeSage was designed to function two ways simultaneously: as an embedded journal inside the funded trading dashboard for prop traders, and as a standalone product for retail traders, trading coaches, and trading groups. The mission was to empower traders to understand their performance, identify behavioral and technical patterns, and evolve into consistently profitable traders through clean data visualization and automated AI insights.
The competitive bet was on combining sophisticated AI analysis with social engagement features and seamless broker integrations — capabilities that point solutions in the market weren't offering together.
My Approach
What I Defined
Advanced Analytics
P&L summaries, win rate, time and instrument-based performance, behavioral analytics, and auto edge detection.
AI Coaching
AI-first interface, daily insights, adaptive AI Coach engine, auto mistake detection, and voice-to-journal.
Psychology
Emotional tagging, morning check-ins, discipline scores, fatigue detection, and psychological performance scoring.
Replay & Backtesting
Session and trade-by-trade replay, strategy comparison, what-if simulations, and prop rule stress testing.
Prop Readiness
Prop rules engine, compliance tracking, violation alerts, multi-prop management, and readiness scoring.
Education & Social
Stage-based education, playbooks, leaderboards, mentor workspaces, and AI-generated review packets.
Outcomes
Key Learning
You can't test a recommendation engine the same way you test a button. I invested heavily in defining what "good" looked like for each AI output before any model code was written. In fintech, the trust gap is the product problem — users must understand and trust AI recommendations before acting on them. That insight shaped every UX and copy decision.
Sample Work
A comprehensive strategic plan for rebuilding, rebranding, and relaunching TradeSage as the leading AI-powered trading journal platform.
Sample Work
A full breakdown of 50+ features across six product areas, prioritized by Must-Have, Should-Have, and Differentiator tiers.
PM Artifacts
Deep-dive documents from the TradeSage buildout.
Roadmap
TradeSage Product Roadmap
4 phases · Q4 2025 – 2027 · AI trading journal platform
Feature Prioritization
Feature Categorization Table
6 product areas · Must-Have / Should-Have / Differentiator
PRD
Journal Dashboard & AI Performance Layer
Core MVP · AI insights · behavioral analytics
BRD
TradeSage Platform Relaunch
Business case · market opportunity · stakeholder requirements
Feature Spec
Adaptive AI Coach
AI coaching engine · behavioral signals · personalized guidance
Sprint Example
Sprint 1 — Discovery & Foundation
Phase 1 · user research · baseline metrics · brand foundation
Roadmap · Fintech · AI Trading Platform
Strategic Context
TradeSage needed to evolve from an embedded analytics layer into a standalone, market-leading AI trading journal. This roadmap outlines the full strategic arc from rebrand through ecosystem expansion — balancing speed to market with product depth.
Phase Overview
| Phase | Timeline | Focus |
|---|---|---|
| Phase 1 · Relaunch & Rebrand | Q4 2025 · 1–2 months | User research, UI/UX overhaul, brand refresh, baseline data foundation |
| Phase 2 · Product Expansion | Q1–Q2 2026 · 3 months | Feature parity, full AI/LLM integration, enhanced analytics, broker integrations, social features |
| Phase 3 · Growth & Monetization | Q3–Q4 2026 | Freemium model, affiliate partnerships, white-label for prop firms, enterprise plans |
| Phase 4 · AI Ecosystem | 2027 | Advanced pattern recognition, predictive analytics, deep platform integrations, strategic partnerships |
Success Metrics
Full Roadmap
A full walkthrough of the TradeSage product strategy — architecture, phases, monetization, and long-term vision.
PRD · Product Requirements Document · Fintech AI
Overview
Funded and retail traders generate significant trade data but lack the tools to extract behavioral and performance insight from it. Existing journaling platforms offer raw data views without meaningful synthesis. TradeSage's Journal Dashboard solves this by combining a clean journaling interface with an AI performance layer that surfaces patterns, flags anomalies, and delivers personalized coaching — all in one place.
Goals & Non-Goals
In Scope
Out of Scope (Phase 2+)
User Personas
Primary — Prop Trader
Uses TradeSage inside the funded platform dashboard. Needs prop rule compliance visibility, daily P&L tracking, and performance pattern identification to protect their funded account.
Secondary — Retail Trader
Standalone subscriber. Wants to improve consistency, understand behavioral patterns behind losses, and receive actionable AI coaching without needing to interpret raw data themselves.
Tertiary — Trading Coach
Monitors student performance. Needs access to trade breakdowns, discipline scores, and behavioral trend data to deliver informed coaching feedback.
User Stories
| ID | User Story | Priority |
|---|---|---|
| US-01 | As a trader, I want to see my daily P&L summary so I can track performance at a glance without manually calculating results. | Must-Have |
| US-02 | As a trader, I want to tag my emotional state on each trade so I can identify how psychology impacts my performance over time. | Must-Have |
| US-03 | As a trader, I want to receive an AI-generated performance note each day so I can understand what my data means without manual analysis. | Must-Have |
| US-04 | As a trader, I want to complete a morning check-in before trading so I can track how my mental state correlates with session outcomes. | Must-Have |
| US-05 | As a trader, I want to see performance broken down by time of day and instrument so I can identify when and what I trade best. | Must-Have |
| US-06 | As a trader, I want a discipline score calculated each session so I can hold myself accountable to my trading rules. | Must-Have |
| US-07 | As a trader, I want weekly and monthly summary reports so I can review progress over longer timeframes without rebuilding data manually. | Should-Have |
| US-08 | As a coach, I want to view a student's discipline score and emotional trend data so I can give informed feedback during sessions. | Should-Have |
Success Metrics
BRD · Business Requirements Document · Fintech
Executive Summary
TradeSage exists as an embedded analytics feature within a funded trading platform. The opportunity is to transform it into a dual-purpose, standalone AI trading journal that competes directly in the trade journal market while retaining its embedded value for prop traders. The relaunch requires a full rebrand, UX overhaul, AI integration, and go-to-market strategy targeting four distinct customer segments.
Business Objectives
Market Opportunity
The trading journal market is growing alongside the rise of retail and prop trading. Existing tools are fragmented — some offer analytics, some offer journaling, none offer a unified AI coaching layer. TradeSage's competitive position is the combination of all three plus social engagement features, giving it a moat that point solutions cannot easily replicate.
Stakeholder Requirements
| Stakeholder | Primary Need | Success Criteria |
|---|---|---|
| Product | Clear scope, phased roadmap, defined success metrics | Phase 1 delivered on time with baseline data established |
| Engineering | Clearly defined requirements, prioritized backlog, stable API contracts | No scope creep mid-sprint; acceptance criteria clear before dev starts |
| Design | Brand direction, design system ownership, user research access | Complete design system delivered before Phase 2 build begins |
| Marketing | Rebranded assets, positioning framework, launch timeline | Website and brand refresh live at Phase 1 close |
| Business | Revenue growth, market share, retention metrics | 100K MAU within 12 months, 70% 90-day retention |
Constraints & Risks
Feature Spec · AI Coaching · Differentiator
What It Is
AI Coach is TradeSage's core AI differentiator — a conversational coaching engine that analyzes a trader's journal data, behavioral signals, and performance patterns to deliver personalized guidance. Unlike static rule engines or generic AI chat, AI Coach adapts its coaching style and recommendations to each user's specific setup, tendencies, and psychological profile over time.
Core Capabilities
Daily Insights
Generates a personalized daily performance note after each session, surfacing patterns the trader may not have noticed.
Auto Mistake Detection
Identifies recurring errors — overtrading after losses, ignoring stop rules, trading outside peak hours — and flags them proactively.
Guardrails
Pre-session rule checks and real-time alerts when a trader's behavior starts drifting from their own defined trading plan.
Voice to Journal
Allows traders to dictate post-trade notes verbally. AI Coach transcribes, categorizes, and links the note to the relevant trade automatically.
AI Unlock Gating
Advanced AI Coach features unlock progressively as a trader builds journal history, incentivizing consistent use and creating a natural upsell path.
Usage-Based Credits
AI analysis depth is governed by a credit system. Free users get daily insights; Standard and Pro users get full coaching access.
Technical Approach
Edge Cases & Risks
Success Metrics
Sprint · Phase 1 · Discovery & Foundation
Sprint Goal
This sprint focused on completing the discovery work required before any design or engineering could begin. The goal was to understand the current user base, define the competitive landscape, and establish baseline metrics — so every subsequent decision would be grounded in real data rather than assumptions.
Team & Ceremonies
Team
Ceremonies
Sprint Planning — Day 1 · 2 Hours
The backlog was groomed in the week prior, so planning focused on commitment and clarity — not estimation or debate. Stories were reviewed, questions were answered, and the team left with a shared goal and a clear first action.
Part 1 · 45 min
What are we building?
Part 2 · 60 min
How are we building it?
Part 3 · 15 min
Alignment and close
User Stories
| ID | Story | SP | Status |
|---|---|---|---|
| S1-01 | Conduct 5 user interviews with current platform traders to surface key pain points with the existing journal experience. | 5 | Done |
| S1-02 | Pull and synthesize 90 days of support ticket data to identify recurring product complaints and feature requests. | 3 | Done |
| S1-03 | Complete competitive audit of 5 trading journal platforms documenting feature gaps and positioning opportunities. | 5 | Done |
| S1-04 | Extract baseline metrics: MAU, DAU, feature usage frequency, and CSV import adoption rates from current platform. | 3 | Done |
| S1-05 | Deliver initial brand direction moodboard and present two visual identity directions to stakeholders for alignment. | 4 | Done |
Definition of Done
Sprint Summary
Feature Prioritization · Fintech · AI Trading Platform
Strategic Context
With a broad product surface spanning analytics, AI coaching, psychology, and education, prioritization was critical. This framework organized 50+ features across six areas into three tiers — what had to ship, what should ship, and what would differentiate TradeSage in a competitive market.
Priority Framework
Must-Have
Core functionality required for launch. Without these, the product doesn't work.
Should-Have
High-value features that round out the experience and drive retention.
Differentiator
Competitive moat features that set TradeSage apart in a crowded market.
Full Feature Table
A complete breakdown of features across Advanced Analytics, Psychology, Education, Replay & Backtesting, AI Coaching, and Prop Readiness.
EdTech · Enterprise Platform · Data-Driven
The Problem
Apex was growing rapidly but training wasn't keeping pace. Knowledge was siloed, onboarding was inconsistent, and there was no systematic way to measure whether training was actually working. New hires took too long to ramp and the company lacked infrastructure to support a growing, distributed team.
My Approach
Outcomes
Key Learning
The principles of good product design and good learning design are nearly identical. Both require understanding the user's goal, removing friction, and measuring whether the experience worked. My background in linguistics and instructional design wasn't separate from the PM work, it was the PM work.
Sample Work
A detailed walkthrough of an AI support training initiative, covering architecture, implementation challenges, and measurable outcomes.
Sample Work
A comprehensive approach to building consistent leadership capabilities from front-line supervisors to executive leadership, including program structure, blended learning methodology, and business impact.
Sample Work
A comprehensive approach to standardizing and optimizing an employee onboarding experience, including AI-powered personalization, cross-functional collaboration, and measurable reduction in onboarding time.
Sample Work
A strategic audit spanning multiple departments — eliminating redundant content, creating role-based learning journeys, and reducing time-to-proficiency.
PM Artifacts
Process artifacts from the Apex LMS buildout — the roadmap, OKRs, and sprint documentation that drove execution.
Roadmap
Training Development Roadmap
Q3 2023 – Q3 2024 · 4 phases · full LMS buildout
OKRs
FY 2024 Objectives & Key Results
3 objectives · 9 key results · outcomes mapped to KPIs
Sprint Example
Sprint 12 — Leadership Track Build
5 user stories · 18 SP · 2-week sprint
Roadmap · EdTech · LMS Platform
Strategic Context
Apex was scaling rapidly with no centralized training infrastructure. This roadmap defined the full buildout from platform selection through optimization — structured in four phases to balance speed of delivery with quality of output.
Roadmap Overview
| Phase | Quarter | Epics & Deliverables | Status |
|---|---|---|---|
|
Phase 1: Foundation Infrastructure & Alignment |
Q3 2023 |
|
Complete |
|
Phase 2: Core Build Launch & First Rollout |
Q4 2023 |
|
Complete |
|
Phase 3: Scale Depth & Breadth |
Q1–Q2 2024 |
|
Complete |
|
Phase 4: Optimize Measurement & Iteration |
Q3 2024 |
|
Complete |
Outcomes by Phase End
OKRs · FY 2024 · Learning & Development
Context
These OKRs were set at the start of FY 2024 to align the learning platform's development directly to business outcomes. Each key result was designed to be measurable, time-bound, and tied to a departmental KPI. Progress was reviewed monthly with department heads and reported to leadership quarterly.
Objective 1
KR 1.1
Launch LMS platform serving 100% of employees by end of Q1 2024
KR 1.2
Achieve 80%+ completion rate on all mandatory training tracks by Q3
KR 1.3
Reduce new hire ramp time significantly from an extended baseline by Q2
Objective 2
KR 2.1
Measurably improve knowledge accuracy scores vs. prior year baseline across key departments
KR 2.2
Deploy adaptive assessments across all core departments by end of Q2
KR 2.3
Achieve 4.2/5.0 average satisfaction score on training programs (quarterly survey)
Objective 3
KR 3.1
Map 100% of training objectives to departmental KPIs across all departments by Q1
KR 3.2
Launch leadership development program covering 100% of managers across all departments by Q2
KR 3.3
Deliver monthly learning analytics report tied to business metrics starting Q2
Sprint Planning · Agile · LMS Build
Sprint Goal
Design, build, and deploy the front-line leadership learning track in TalentLMS — fully tested, assigned to a pilot cohort of managers, and ready for structured feedback collection.
Team & Ceremonies
Team
Ceremonies
User Stories
As a front-line manager, I want a defined learning path so I know exactly what to complete to develop my leadership skills.
Acceptance Criteria
As a training admin, I want to assign the leadership track to specific user groups so I can control access and track completion at scale.
Acceptance Criteria
As a department head, I want to see my team's completion rates so I can track progress toward our Q2 leadership development goals.
Acceptance Criteria
As a learner, I want a knowledge check at the end of each module so I can verify my understanding before progressing to the next.
Acceptance Criteria
As the PM, I want 25 managers enrolled in the pilot track so I can gather structured feedback before full organizational rollout.
Acceptance Criteria
Definition of Done
Sprint Summary
Consumer · SaaS · AI · Social Platform
The Problem
Grandma's handwritten recipe cards get lost. A family's secret sauce nearly dies with a generation. Diaspora communities lose their food culture one generation at a time. Existing apps treat food as content, exposing personal recipes in algorithm-driven feeds designed for strangers, not families. Nobody owned the private, intimate space where families preserve and grow their food heritage together.
The Vision
Griffin's Kitchen is a private, invitation-based social recipe platform. Families and cultural communities create shared "pods", living digital cookbooks that grow richer over time. Members add recipes, attach personal stories and memories, tag contributors, and trace the origin of every dish. The AI layer surfaces suggestions based on available ingredients and dietary needs. The ultimate upsell: a beautifully designed, professionally printed hardcover cookbook generated from any pod's collection.
Product Strategy
Market Opportunity
PM Thinking
The most important early product decision was committing to a private, invitation-based model. Privacy is the core value proposition, it's what makes users comfortable sharing a grandmother's handwritten card or a dish with deep personal meaning. Private-first means slower top-of-funnel growth but dramatically higher retention and willingness to pay. The physical book upsell isn't a forced add-on, it's the natural culmination of everything a pod builds together.
PM Artifacts
The following artifacts were developed from the Griffin's Kitchen business plan to illustrate PM deliverable formats.
PRD
Pod Creation & Invite System
Core MVP feature · Week 1–2 · private invite model
Roadmap
12-Week MVP Roadmap
6 phases · Foundation through Growth
Feature Spec
Printed Cookbook Export
Print-on-demand · Lulu xPress API · $45–65
OKRs
Year 1 Growth OKRs
Revenue milestones · retention · community growth
PRD · Product Requirements Document
Overview
Families have no private, shared digital space to collect and preserve recipes together. Existing platforms are built for public audiences, not intimate family groups. Griffin's Kitchen solves this through "pods" — private, invitation-only spaces where families and communities build living cookbooks together.
The Pod Creation & Invite System is the foundation of the entire product. Without it, nothing else is possible. It must be simple enough that a grandparent can set it up, and private enough that a family trusts it with personal recipes.
Goals & Non-Goals
Goals
Non-Goals (v1)
User Stories
Pod Creator
"As a family organizer, I want to create a private pod for my family so we can collect recipes in one place without strangers seeing them."
Acceptance: Pod created in <3 steps; default visibility is Private; creator is auto-assigned Admin role.
Inviter
"As a pod Admin, I want to invite family members by email so they can join without needing to know my account details."
Acceptance: Invite email sent within 60s; link expires in 7 days; invitee lands on pod onboarding flow.
Invitee
"As someone who received an invite, I want to join my family's pod without confusion so I can start contributing recipes right away."
Acceptance: Account creation + pod join in <2 minutes; first recipe prompt shown on successful join.
Functional Requirements
| ID | Requirement | Priority | Notes |
|---|---|---|---|
| FR-01 | User can create a pod with name, description, and optional photo | P0 | Name required; max 50 chars |
| FR-02 | All pods are private by default; no public discovery | P0 | Core privacy promise |
| FR-03 | Admin can invite members via email address | P0 | Transactional email via Resend |
| FR-04 | Admin can generate a shareable invite link (7-day expiry) | P1 | Revocable by Admin |
| FR-05 | Roles: Admin (full), Contributor (add/edit own recipes), Viewer (read-only) | P0 | Stored in Supabase RLS |
| FR-06 | Member cap enforced by tier: Free (3), Family Pod (10) | P0 | Upsell prompt at limit |
| FR-07 | Admin can remove members and transfer Admin role | P1 | |
| FR-08 | New member sees a "first recipe" prompt immediately after joining | P1 | Activation moment |
Technical Approach
pod_invites table with expiry timestamp/invite/[token] route that validates token before prompting account creationSuccess Metrics
Open Questions
Roadmap · 12-Week MVP Plan
Strategic Context
The roadmap is sequenced to validate the core privacy-first thesis before investing in AI features or monetization. Phases 1–2 prove the product works. Phase 3 makes it delightful. Phases 4–5 make it revenue-generating. Phase 6 makes it grow. Nothing in Phase 3 onward gets built until Phase 1–2 user testing confirms the pod model resonates.
MVP Phases
| Phase | Weeks | Focus | Key Deliverables |
|---|---|---|---|
| Phase 1 | 1–2 | Foundation — Auth, pod model, data schema | Supabase auth + RLS; pod creation; invite system; member roles; basic recipe CRUD |
| Phase 2 | 3–5 | Core Product — Recipe experience and social layer | Rich recipe editor (ingredients, steps, photos, story notes); contributor attribution; pod feed; recipe tagging; search within pod |
| Phase 3 | 6–7 | AI + Polish — Intelligent suggestions and UX refinement | Claude API integration; ingredient-based meal suggestions; dietary substitution assistant; onboarding flow polish; mobile responsiveness |
| Phase 4 | 8–9 | Monetization — Stripe subscriptions and tier enforcement | Stripe integration; Free / Individual / Family Pod tier enforcement; upgrade flows; billing portal; usage-based upsell prompts |
| Phase 5 | 10–11 | Print + Launch — Cookbook export and public launch prep | React-PDF cookbook generator; Lulu xPress API integration; order flow; payment capture; beta program; marketing site; waitlist |
| Phase 6 | 12 | Growth — Referral mechanics and community launch | Pod referral system; community launch (diaspora groups, family associations); PostHog analytics; NPS baseline; product-led growth loops |
Post-MVP Priorities
Community Tier
Larger pods (50+ members) for cultural organizations, churches, and community groups — custom pricing
Heritage Digitization
Photo-to-recipe OCR for scanning handwritten cards — AI-assisted transcription and formatting
Recipe Versioning
Track edits and variations over time — "grandma's original" vs "how mom makes it now"
Mobile App
Native iOS/Android — in-kitchen mode, voice recipe input, step-by-step cooking mode
Revenue Logic
Each phase is sequenced to unlock the next revenue layer. The freemium tier fills the top of funnel. Paid subscriptions convert the most engaged pods. The printed cookbook captures high-margin one-time revenue at the point of peak emotional investment — when a family has built something they want to hold in their hands. The sequence is designed so that each phase funds the next and reduces the capital required to reach sustainability.
Feature Spec · Printed Cookbook Export
Why This Feature
A printed cookbook is the highest-value expression of everything a pod has built together. It transforms digital memories into a physical heirloom. From a business perspective, it's the highest-margin transaction in the product — no recurring cost, no support overhead after print fulfillment, and strong emotional pull that drives the purchase decision independent of price sensitivity. It also serves as organic word-of-mouth: a printed book on a coffee table becomes a conversation and a referral.
User Flow
Trigger: Admin clicks "Print Our Cookbook" from pod settings
CTA appears once pod has ≥10 recipes. Tooltip explains what the feature does.
Recipe Selection: choose which recipes to include
Defaults to all pod recipes; Admin can deselect; sorted by category or contributor.
Customize: choose cover, title, dedication, and section headers
3 cover templates; custom title (defaults to pod name); optional dedication page.
Preview: live PDF preview rendered in-browser
React-PDF renders a paginated preview; Admin can scroll and adjust before ordering.
Order: enter quantity, shipping address, and complete payment via Stripe
Pricing: $45 (softcover, 1 copy), $65 (hardcover, 1 copy). Bulk discounts at 3+ copies.
Fulfillment: PDF sent to Lulu xPress; printed and shipped directly to customer
Est. 7–14 business days. Tracking number emailed on dispatch. Griffin's Kitchen handles no physical inventory.
Technical Spec
PDF Generation
@react-pdf/renderer)Print Fulfillment
Economics
The printed cookbook is the highest-margin product in the suite. Pricing is set at a premium that reflects the emotional value of the artifact — a family heirloom — not just the cost of production. Print-on-demand fulfillment via a third-party API means zero inventory risk and no physical operations overhead. Gross margin after print and payment processing is healthy across both formats, making this the most capital-efficient revenue line in the business.
Edge Cases & Risks
OKRs · Year 1 Growth Objectives
Framework
Griffin's Kitchen measures success by pod health, not just user count. A pod with one person adding recipes is a failed pod — the value only exists when families are building together. The north star metric is active multi-contributor pods: pods with 3+ contributors who have each added at least one recipe in the last 30 days. Revenue and growth KRs are downstream of that.
Q1 OKRs
Objective
Launch a product families love and want to share with each other.
Hit Q1 revenue target; subscription growth on plan
50+ active pods with ≥3 contributors each
≥60% of pod creators invite at least 2 members within their first 7 days
Average pod has ≥5 recipes added within first 30 days of creation
Q2 OKRs
Objective
Make Griffin's Kitchen the place families return to every week, not just when they remember.
Hit Q2 revenue target; growth rate sustaining month-over-month
Weekly active pod engagement rate ≥40% (at least one recipe action per week)
AI meal suggestion feature used by ≥30% of active contributors in any given week
Month 6 churn below 5% for paying pods
First 25 printed cookbooks ordered
H2 OKRs
Objective
Prove the business model is self-sustaining and position for acquisition or Series A.
Reach 12-month revenue target; business operationally self-sustaining
500+ active multi-contributor pods
Printed cookbook revenue represents ≥20% of total revenue
12-month net revenue retention ≥110% (expansion from Family → Community tier upgrades)
NPS ≥50 from paying pod admins
Key Risks
Activation Risk
Pod creators don't invite enough members. Mitigation: in-app prompts at key moments, email nudges after 48 hours of solo activity.
Churn Risk
Seasonal engagement (holiday spikes, summer dips). Mitigation: AI-driven recipe discovery keeps pods active year-round.
Revenue Mix Risk
Print orders underwhelm. Mitigation: physical cookbook positioned as milestone gift — target holiday and lifecycle moments (weddings, reunions).
E-Commerce · Web Development · Publishing
The Brief
Camille Pandian Milner was preparing to launch her debut novel, Climbing Up the Walls, a character-driven story set in London's mid-2000s music scene. She needed more than a basic website. She needed a platform that could tell her story as an author, build anticipation ahead of publication, sell directly to readers, and capture an audience through email, all in one cohesive experience.
What I Built
I designed and developed the complete site on Shopify, building a custom experience that goes well beyond a standard storefront. The platform includes:
Discovery
Before touching Shopify, I sat down with Camille to understand what she was actually trying to create. We talked about the book — the themes, the atmosphere, the specific world of London's music scene in the early 2000s. Late nights, underground venues, the energy of that era. She wasn't just launching a product. She was inviting people into a world she'd spent years building.
From that conversation, the brief became clear: the website needed to feel like the book. It had to carry the same atmosphere. And it needed to function as a real business tool — a store, an email list, a press page — not just a pretty landing page.
Design Decisions
The dark background wasn't a trend choice — it was a deliberate reference to the late-night world the book inhabits. The cyan and coral accent palette evokes the neon and warmth of that music era without being literal about it. Serif headlines (Newsreader) give the site literary weight; the clean sans-serif body text (Archivo) keeps it modern and readable.
The photo journal section was a key decision. Rather than just showing the book cover, I built an immersive, zoomable gallery of archival images from the London music scene that inspired the novel. This wasn't decorative — it was designed to make a visitor feel something before they ever clicked "buy."
Product Thinking
Most author websites are static — they exist to confirm that a book exists. I approached this differently. The site was designed with a clear user journey in mind: arrive, get emotionally invested in the story, join the waitlist, and stay connected through launch and beyond.
Every section serves a conversion goal. The photo journal builds atmosphere and keeps people on the page. The multiple format options let readers self-select by preference. The waitlist captures intent before inventory exists. The mailing list turns one-time visitors into a community the author owns — not one that lives inside an algorithm.
Process
The project ran over about a month, with multiple rounds of revisions. Camille reviewed, gave feedback, and I adjusted. This back-and-forth wasn't friction — it was part of the process. A website for a debut novel needs to feel personal, and getting there required real collaboration. We kept iterating until the site matched the vision she'd carried through years of writing the book.
Live Site
The finished platform, live at unboundbooks.net.
Outcomes
AI · EdTech · B2B SaaS · Venture Concept
The Problem
Average onboarding takes 3–6 months. 70% of employees report inconsistent training experiences across departments, leading to knowledge gaps and performance variation. HR and L&D teams spend the majority of their time on repetitive, low-leverage tasks instead of strategic initiatives. The result: significant cost per new hire from inefficient onboarding, and early attrition that compounds it. Companies invest heavily in training content that sits unused because it's static, hard to find, and disconnected from how people actually work.
The Vision
Accendra is a multi-agent platform designed to transform how organizations onboard, upskill, and retain talent. Rather than a single chatbot or LMS add-on, it's a coordinated system of specialized agents — each owning a distinct part of the training workflow — that adapt to each employee's role, pace, and knowledge state.
Onboarding Agent
Creates personalized learning paths based on role, experience, and departmental needs. Adapts in real time to learning pace and progress signals.
FAQ & SOP Agent
Provides instant, contextually accurate answers from existing documentation. Eliminates the knowledge-retrieval bottleneck that stalls new hires daily.
Assessment Agent
Conducts adaptive evaluations, identifies skill gaps, and adjusts learning paths dynamically. Surfaces insights to managers without requiring manual tracking.
Curriculum Design Agent
Automates instructional design by analyzing org needs, industry standards, and learning objectives — turning weeks of L&D work into hours.
Market Opportunity
The corporate training market is growing at 20% CAGR. The convergence of remote work, skills shortages, and AI adoption creates compounding demand for automated, personalized training infrastructure. Target buyers: HR departments, L&D teams, SaaS companies, fintech firms, and enterprise operations teams scaling rapidly.
Competitive Edge
| Competitor Type | What They Do | Their Gap |
|---|---|---|
| LMS Platforms | Deliver standard training content | No personalization, static content, poor engagement |
| HR Chatbots | Answer basic FAQs | Not training-focused, no learning context or progression |
| Knowledge Bases | Store SOPs and documentation | Passive consumption only, no adaptive learning |
Why I'm Building This
Accendra isn't a research project — it's the product I wish I'd had. I spent years inside organizations where training was the bottleneck: knowledge siloed in people's heads, new hires floundering for months, L&D teams buried in repetitive requests. I built manual versions of what Accendra automates. The agent layer is what makes it scalable.
My background in applied linguistics, instructional design, software engineering, and AI product management puts me in a rare position to design a system that's both pedagogically sound and technically viable. That combination is the moat.
Full Concept Deck
A full walkthrough of the Accendra vision — market opportunity, agent architecture, business model, competitive landscape, and growth strategy.