Go to home page
BACK TO HOME
PRODUCT ENGINEERING CASE // ACTIVE

CRAMAI

Multi-agent AI learning assistant

OPEN PRODUCT

FULL PRODUCT DEMO

CHINESE · 01:05

PRODUCT LOOP

01
Bring materials
PDF · VIDEO · LINK
02
Learn with Tutor
READ · ASK · CREATE
03
Practice & test
QUIZ · MOCK EXAM
04
Track & improve
REPORT · NEXT PLAN
SHIPPED_OUTCOME

I helped move the product from beta-code access toward open trials, then built guest exams, AI Tutor learning features, a structured item bank, and a commercialization feedback loop.

MY ROLE
Product Lead / Product Engineer
Dec 2025 - Present
OPEN CASE OVERVIEW

CramAI turns PDFs, videos, and links into a conversational, testable, trackable learning loop. I own product direction, priorities, and feedback review while contributing product design, UI/UX, frontend, selected backend work, and analytics.

A 65-second Chinese walkthrough of CramAI's entry point and core learning experience. An English edition will be added later.

TEAM & OWNERSHIP

TEAM SIZE
Cross-functional team, peaking at 9
Product Lead / Product Engineer

I OWNED

Product direction, priorities, and release sequencing
Reviewing market/user feedback and turning it into product decisions
Product development, UI/UX, frontend, and recent selected backend work
Commercialization design, event instrumentation, and GA / Plausible funnels

COLLABORATION BOUNDARY

AI-agent backend and algorithms were owned by other team members
Core multimodal RAG capabilities were owned by other team members
AWS infrastructure and deployment were owned by other team members
OPEN ROLE DEFINITION

As Product Lead / Product Engineer, I own direction, priorities, feedback review, and the main product experience while contributing UI/UX, frontend, recent selected backend work, and commercialization analytics. Algorithms, multimodal RAG, the AI-agent backend, and AWS deployment are owned by other team members.

REAL USAGE SIGNALS

UPDATED // 2026-09-05
SIG_01
1,264
ACTIVE USERS IN 90 DAYS
SIG_02
22,514
PRODUCT EVENTS
SIG_03
2M 29S
AVG ENGAGEMENT PER ACTIVE USER
SIG_04
1,670
ZH + EN DASHBOARD VIEWS
LOADING_DATA_STREAM...
OPEN METHODOLOGY & DATA NOTE

Early production data shows this is a product people use, not a concept left at prototype stage: users reached the landing, learning workspace, and exam paths.

Period: Jun 7–Sep 4, 2026. These are aggregate GA4 usage signals, not claims of growth, retention, or conversion lift. Funnel conclusions will be updated after event-contract and internal-traffic cleanup.

PRODUCT SYSTEMS

04 MODULES
SYS_01

ACCESS & GROWTH LOOP

Moving between beta, open trials, and registration conversion

CramAI English production landing page
LIVE_PRODUCTION // PUBLIC EXPERIENCE ENTRY
KEY DECISION
→
TRADEOFF
→
SHIPPED OUTCOME

Access strategy can change with the product stage. Guests experience questions, grading, and reports before authentication becomes the natural next step.

OPEN DECISION LOG

The product moved through beta-code access, open registration, and renewed research-led gating. I owned the access strategy for each phase and made guest exams the clearest path to understanding the core value.

KEY DECISION

Centralize access rules and put value before authentication: guests choose Quick 12 or Full 180, complete an exam, then register around high-intent moments such as results and follow-up plans.

TRADEOFF

Guest access increases cost and abuse risk, so the trial is complete but controlled while account-level and high-cost capabilities remain gated.

  • +Managed beta-code access, open registration, and research-led gating across product stages
  • +Designed and shipped Quick 12 / Full 180 guest CFA exam modes
  • +Replaced full-page login redirects with a progressive, email-first auth modal
  • +Designed contextual paywalls and trials around the user's current intent
SYS_02

AI TUTOR LEARNING SYSTEM

Bringing the FeaturePanel toolkit into the conversation

CramAI English AI Tutor learning conversation
LIVE_PRODUCTION // AI TUTOR WORKSPACE
KEY DECISION
→
TRADEOFF
→
SHIPPED OUTCOME

Users can read, discuss, summarize, test, and review the same material in one flow; FeaturePanel becomes a learning system rather than a toolbox.

OPEN DECISION LOG

FeaturePanel combines summaries, mind maps, flashcards, quizzes, knowledge posters, podcasts, resources, and notes. I set the product mix, interaction model, and priorities, then integrated those outputs into the AI Tutor conversation and active-material context.

KEY DECISION

Do not present generation tools as isolated buttons. Center the active material and Tutor conversation so users move continuously between reading, asking, and generating learning outputs.

TRADEOFF

A growing toolkit increases cognitive load, requiring shared entry, state, and output patterns plus device-specific material navigation.

  • +Planned and prioritized summaries, mind maps, flashcards, quizzes, posters, podcasts, and more
  • +Connected FeaturePanel outputs to the AI Tutor conversation and active material
  • +Rebuilt DocumentViewer multi-file switching and mobile behavior
  • +Improved PDF and video resume behavior for long-form study
SYS_03

EXAM & ITEM BANK

From demo papers to an operable item bank

CramAI English full-screen mock exam interface
LIVE_PRODUCTION // FULL-SCREEN EXAM
KEY DECISION
→
TRADEOFF
→
SHIPPED OUTCOME

Exam content can enter the product through a controlled workflow and support answering, grading, weakness diagnosis, error review, and post-exam planning.

OPEN DECISION LOG

Quick 12 / Full 180 validated the exam experience, but demo data could not support ongoing operations. My recent focus is a structured item bank and preview-confirm import flow so formal exam content can be safely reused, traced, and updated.

KEY DECISION

Model raw imports, papers, paper items, and questions separately, with a preview-and-confirm step before operational content is written.

TRADEOFF

A structured model is more complex than storing one JSON blob, but it enables reuse, version history, import replay, and future expansion.

  • +Wrote the phase-1 PRD and schema for the item bank with paper/question separation
  • +Built a guarded preview-and-confirm import workflow in the admin panel
  • +Shipped the full-screen answering, grading, radar chart, and post-exam plan loop
  • +Turned the error book from MVP into a flag-gated product feature
SYS_04

COMMERCIALIZATION & FEEDBACK

Using GA and Plausible to test product paths

CramAI English subscription plans and commercialization interface
LIVE_PRODUCTION // SUBSCRIPTION VALIDATION
KEY DECISION
→
TRADEOFF
→
SHIPPED OUTCOME

Registration and activation events form one continuous journey, allowing the team to locate drop-off instead of relying on isolated page views.

OPEN DECISION LOG

I contributed commercialization design and owned key event instrumentation, registration funnels, and attribution fixes. The data will not be dressed up as vanity metrics; it is used to find drop-off and decide which product assumption to revisit.

KEY DECISION

Define one deduplicated event contract from landing and guest experience through authentication, activation, and first message, with client/server identity alignment.

TRADEOFF

A shared contract and identity alignment take longer than ad-hoc tracking, but prevent duplicate conversions and broken journeys across login.

  • +Defined the registration funnel event contract from landing to first message
  • +Fixed lost production conversion events and user identity alignment
  • +Ensured balance refresh failures never lose data and guest sessions never block the main entry

KEY REFLECTIONS

REFLECTION_01

Access is not a one-time switch. Beta codes, open registration, and research-led gating must move with the product stage instead of becoming permanent page logic.

REFLECTION_02

More features do not automatically create value. FeaturePanel only becomes a learning loop when it shares the AI Tutor conversation and material context.

REFLECTION_03

Demo papers validate experience, not content operations. Quick / Full modes exposed the need for a structured item bank and controlled import.

REFLECTION_04

Data does not need to look good; it needs to be trustworthy. Funnel evidence exists to expose problems and inform the next priority.

TECH STACK

INTERFACE
Next.js 15React 19TypeScriptTailwind CSSHeroUI
PRODUCT API
Hono (Node.js)FastAPI (Python)Prisma
AI SYSTEM
LangGraphLlamaIndexAWS Bedrock
MEASUREMENT
EChartsGA4Plausible