Products / Life · Work · Research

AI Native. Creative Team.

Personal reflection. Informed decisions.

We think with AI and create new experiences through creativity. From personal reflection with OpenAlarm to portfolio-based research with OpenStockAI, explore how our ideas become products.

01 / Available now

OpenAlarm

Your patterns. Your feelings. A reply that connects.

OpenAlarm considers recurring patterns and emotional context in your diary to shape a personal AI reply. Explore suggestions based on alarm use, revisit past entries and trace place-linked plans on a footprint map.

Alarm patterns and suggestions
Pattern- and emotion-aware diary replies
Place-linked footprints

Visit OpenAlarm

Protecting personal-record inputs and outputs

Records are handled within the user's account scope. Diary input is separated from instructions, and replies containing secrets or internal information are rejected before storage with a security alert.

Explore the protection flow

Everyday lifeOpenAlarm workflow →

02 / Available now

OpenStockAI

Research that starts with your holdings.

OpenStockAI puts a question in portfolio context, routes it through the appropriate research path, and organizes the response around evidence, dates and risk. It is research support, not a return promise.

Portfolio context
Routed agent work
Sources and review

Visit OpenStockAI

Security checks at request time

Research agents work alongside a security team. Checks identify exfiltration and privilege-bypass signals, connect request blocking to redacted audit records, and support review.

Explore the protection flow

Investment researchOpenStockAI workflow →

Orchestration & protection

AI that divides the work. Boundaries that stay in place.

Orchestrators connect tools and agents, while product-specific safeguards check access, inputs and outputs. Explore where detection, blocking and limits on sensitive-data transmission take place.

A Privacy Harness checks task-specific allowlists before information reaches an external model. It limits unnecessary source and identifying data and blocks policy-violating requests.

OpenDataLabs / 01

Input

Documents · handwriting · extracted candidates

Selected work

OCR & extraction tools
Allowed model assistance
Rules & human review

Privacy Harness before external AI

AI input protection and training policy

Default assistance uses minimal structured information. Repair tasks that send short text fragments and separately enabled document-image analysis follow their task-specific allow policies.

Read the protection scope

OpenDataLabs is in development. The harness is a technical control on transmission; model-provider retention and training policies require separate verification.

A harness applies safeguards around AI execution. Transmission controls, redaction and output checks reduce exposure of sensitive information; provider retention and training policies are managed separately.

Inside the work / 03

Behind every result, a process.

Choose a product and follow its process. See what each stage receives, what it does and what it produces.

OpenAlarm / Alarm patterns

From everyday patterns to a considered suggestion.

Alarm records and plans inform specialist analysis, followed by checks on each suggestion.

Workflow walkthrough · On narrow screens, scroll horizontally to explore.

Workflow walkthrough 01 — 04

OpenAlarm

Gather records

Input: Alarm events · plans
Work: Organize usage statistics and previous suggestions.
Output: Context for analysis

Connected stages

Only relevant analysis runs. Automatic application depends on operating settings and the kind of suggestion.

OpenAlarm / Alarm patterns

Alarm records and plans inform specialist analysis, followed by checks on each suggestion.

Gather records
Organize usage statistics and previous suggestions.
Sleep & waking
Examine recurring patterns in alarm use.
Plans & travel
Consider reminders in the context of plans and travel.
Automation
Consider changes that may suit automation.
Check proposals
Check evidence, conflicts and previous rejections. Failed proposals need correction.
Await review
Save suggestions for the user to inspect and act on.
Eligible changes
Apply only eligible changes. Failed applications remain pending.

OpenAlarm / Diary & reply

Consider recurring patterns and emotional context across today's diary and recent entries to shape an AI reply for your situation.

Write
Save a diary entry for its date.
Patterns & emotions
Consider today's diary alongside recent entries to understand recurring user patterns and emotional context.
Personal reply
Generate and save a reply shaped by your patterns, feelings and situation.
Revisit
Revisit entries through the diary, calendar and collection.

OpenAlarm / Footprints

Map the places attached to plans and distinguish visit markers using completion status and location events.

Place-linked plan
Gather plans with a saved location.
Check status
Distinguish visits using completion or location reminder events.
Footprint map
Revisit plans and places on a yearly map.
Diary collection
Browse past diary entries and replies on the same page.

OpenDataLabs / Documents & schedules

Connect outbound checks, OCR re-reading, Jev verification, reference comparison and human review. Bring shipping, delivery and payment deadlines from documents and notes into a schedule with source evidence and completion tracking.

Document input
Read content and source positions using the appropriate parser.
OCR & re-reading
Retry low-confidence reads with preprocessing profiles and compare measured results.
Extract & organize
Extract and normalize fields, marking uncertain values.
Privacy Harness
Validate allowlisted fields, lengths and forbidden keys/values. Block violating requests; exclude digit-containing fragments from text-repair tasks.
Jev verification
When external AI is allowed and connected, use providers such as Jev to verify word and spacing candidates.
Correction checks
Apply internal checks to correction proposals and preserve transformation history.
Compare records
Resolve document fields against reference records.
Check rules
Check quantities, amounts and required fields.
Result & evidence
Keep the machine decision, evidence and review items.
Human review
A person corrects or confirms the fields with a recorded history.
Record confirmation
Preserve the original machine decision separately from human confirmation.
Connect schedules
Collect shipping, delivery and payment deadlines into source-linked events. Leave ambiguous dates unresolved for review.
Track & follow up
Review approaching and overdue events, mark them handled or reopen them. Delivery and shipping events also use linked receipt evidence.

OpenStockAI / Deep team research

Follow the coded team research path: seven specialist roles, department reports, decision analysis, conditional debate and audit.

Question & holdings
Read the question in portfolio context.
Intent & routing
Choose the appropriate response or team research path.
Research agenda
Set the subject and agenda for the team.
Deep financials
Collect financial evidence for the research.
Lighter response
Use a separate response path. Playback below illustrates deeper team research.
Company analysis
Use the specialist's tools and analysis to prepare an opinion. Concurrency is bounded.
Market research
Use the specialist's tools and analysis to prepare an opinion. Concurrency is bounded.
Competitors
Use the specialist's tools and analysis to prepare an opinion. Concurrency is bounded.
Valuation
Use the specialist's tools and analysis to prepare an opinion. Concurrency is bounded.
Future value
Use the specialist's tools and analysis to prepare an opinion. Concurrency is bounded.
Risk
Use the specialist's tools and analysis to prepare an opinion. Concurrency is bounded.
Performance
Use the specialist's tools and analysis to prepare an opinion. Concurrency is bounded.
Department reports
Synthesize opinions by department, handling missing evidence and failures.
Hybrid research
Enrich context through the conditional hybrid research stage.
Decision analysis
Run strategy, trading, fund and portfolio analysis, then synthesize department reports.
Bull case
For eligible single-symbol research, examine the bull case.
Bear case
Examine the bear case in parallel with the bull case.
Debate judgment
Compare arguments and record unresolved conflicts. Fast, comparison and portfolio reviews skip debate.
CIO synthesis
Combine research and discussion into the final synthesis.
Approval gate
When enabled and sensitive proposals exist, await administrator approval. This does not mean trade execution.
Quality & risk audit
Audit quality and risk. Fast and deep modes use different review methods.
Revise synthesis
Revise the synthesis when needed and run another audit.
Report & evidence
Format the report with evidence, dates, risks and limitations.

In the works

OpenDataLabs

Connect verified documents and handwriting to shipping, delivery and payment schedules. Track deadlines and completion with source evidence and human review. In development.

OpenDataLabs workflow →

Privacy Harness before external AI

A Privacy Harness checks task-specific allowlists before information reaches an external model. It limits unnecessary source and identifying data and blocks policy-violating requests.

Explore the protection flow

In development

OpenConnect

A new product in the works.

Planned

사주팀장

A new product in the works.

Planned