swipe to navigate
NeuralRecall.ai · Core

My Digital

Twin.

Core · An AI that learns how you think, never forgets what you know,
and connects the dots across your life — growing with you every day.
01 / 34
Core is another version of you that remembers everything, sees every connection, never loses context and keeps getting smarter. It’s exactly you — just slightly better.
Every professional will have a digital twin.
The question isn’t if.
It’s whose.

Every AI starts
with amnesia.

Monday
Explain project
Tuesday
Explain project
Wednesday
Explain project
Thursday
Explain project
Your AI forgets. You don't.
02 / 16

Your life is scattered
across 14 applications.

None share context.
Gmail
Slack
WhatsApp
Calendar
Drive
Notion
Outlook
Zoom
Teams
Dropbox
LinkedIn
Spotify
GitHub
Health
03 / 16
The knowledge graph

Core connects everything.

Every email. Every meeting. Every document.
Every relationship. One living graph.

42 goals · 106 entities · 182 agents
8 of 13 sources connected.
04 / 16
For the first time, AI understands your context.
05 / 16
Deep context · Entity clusters

People. Projects.
Messages. Connected.

Every cluster contains the full story.
People you know. Messages you sent.
Events you attended. Documents you wrote.

Expand any node. See the source.
Everything traceable. Nothing lost.
06 / 16
This is the moment.

Not search.
Understanding.

Query
"I have a meeting with the Legalfab investors in 20 minutes. What's my strongest angle?"
Core · Thought for 32s
Your strongest angle isn't the tech—it's the velocity. The evidence is in your Legalfab_SHA_Draft, your Gmail from Tuesday, and 4 hours of Deep Focus Spotify while editing the Project_Genesis deck.
07 / 16
Source panel · Full provenance

Every claim linked
to its source.

Beta access requests. Customer questions.
Onboarding flows. Service appointments.

Every message. Every sender.
Every connection. Visible and verifiable.
08 / 16

Your knowledge never
leaves your device.

Local AI.
Private graph.
Verifiable sources.
09 / 16
Scenario 0 · Enterprise · Series D
AeroDynamics Robotics
The pitch has been rehearsed
The numbers have been checked
The story is polished
The lead partner loves the company
Everything points to a successful raise
$900M
Target valuation · Series D · Silicon Valley growth fund
Ten days before the roadshow  ·  6:42am  ·  CEO opens Core

Core worked
through the night.

High Priority · External Intelligence
At 02:17am, Core indexed a podcast released hours earlier. The guest was the largest LP backing the VC leading the Series D. Most people ignored it. Core didn't.
Sources monitored overnight
Email
Meetings
Slack
Jira
Google Drive
GitHub
CRM
Board papers
News
Research
Podcasts
Patents
Podcast · Indexed 02:17am
▶ 47:13
"We're no longer interested in autonomous hardware. We believe the real winners will be companies building the operating systems that coordinate fleets of autonomous machines."
Largest LP · Backing the Series D lead fund
Core marks statement as strategically significant.
Begins connecting dots across 18 months of company history.
Then something extraordinary happens.
Customer workshop
DHL CTO
"Managing one thousand drones will eventually become harder than building one thousand drones."
14 Jan
Teams meeting
Head of Engineering
"Our orchestration software is becoming more valuable than the aircraft themselves."
03 Mar
Draft patent
Filed 5 months ago
Never in a board meeting
Distributed Autonomous Fleet Coordination Engine
5mo ago
CRM notes
3 customers
Independently
Fleet orchestration APIs. Not drones. Software.
2 weeks ago
Core · Causal inference · 18 months · 72 evidence points
The company has been presenting itself incorrectly.
For eighteen months.
Yesterday · Investor deck
Autonomous Drone Manufacturer
→
6:43am · Core rewrites the slide
The Operating System for Autonomous Fleets
"We came looking at a robotics company. We left believing we were investing in the infrastructure layer for autonomous mobility."
Lead partner · Three weeks later · Final partner meeting
Not because someone searched harder.
Not because someone worked longer.
Because Core continuously connected thousands of unrelated moments across time
until a completely different truth emerged.
Core doesn't answer questions.
Core changes the questions worth asking.
Scenario A · Personal · Executive

40 minutes lost
before a board meeting.

Maya Chen — VP Product, Series B startup (80 people) Monday morning. Board meeting at 10am. Maya spent 40 minutes searching Gmail and Slack trying to reconstruct the history of a pricing decision. She opened Core. She never did that again.
Sources connected
Gmail · 12,400 emails
Google Calendar · 890 events
Google Drive · 1,200 files
Slack · 3,100 messages
Local · 340 files
What changed by Week 2
Core detected her board prep pattern after 3 meetings. Now every evening before a board meeting, a brief arrives automatically — unanswered board emails, latest ARR, Slack risks from #leadership. She prepares in 4 minutes, not 40.
Day 1 · Monday 09:12am · board meeting in 48 minutes
Maya asked
"What has our team discussed about the pricing redesign in the last month?"
Core · Thought for 18s
Topic cluster "pricing redesign" → 23 connected messages + documents + events
Slack #product, May 3 — Sarah Kim proposal · 23 replies over 3 days
Gmail May 8–12 — Maya ↔ CFO James Liu · margin impact debate · 4 emails
Drive "Pricing Model Comparison v3" — last edited May 15 · comments from 3 people
Gap: May 14 working group — no documented outcome found in any source
Core — answer
Sarah Kim raised the proposal in #product on May 3 (usage-based model, 23 replies). CFO James Liu flagged in email that usage-based would reduce revenue from your top 10 accounts by ~15%. Option B in Sarah's doc is a hybrid model preserving 92% of current revenue.

Critical gap: The May 14 working group had no documented outcome. Sarah's last Slack message May 16 asked "are we going with Option B?" — no reply found in any source.
Evidence trail
Slack · #product
Sarah Kim — usage-based proposal · 23 replies
03 May
Gmail · CFO
Usage-based reduces top-10 revenue by ~15%
08–12 May
Drive · Sarah
Pricing Model Comparison v3 · Option B hybrid · 92% revenue
15 May
Slack · Sarah
"are we going with Option B?" — unanswered
16 May
Scenario B · Teams · Professional Services

18 months of client
history. 2 hours away.

David Okafor — Freelance management consultant · 6–8 active clients David has a call with Meridian Healthcare in 2 hours. He took them on 18 months ago, handed off to a subcontractor for 6 months, and is now re-engaging. He barely remembers the details. He has 25,000 emails and 430 local documents.
Sources connected · 5 years of consulting
Outlook · 25,000 emails
Outlook Calendar · 2,100 events
OneDrive · 680 files
Local · 430 Word · Excel · PDF
Cross-client intelligence surfaced · Month 2
Core spotted Greenfield Energy nearing Phase 1 completion — no Phase 2 discussion in any source. David's own SOW patterns show 40% average growth at this stage. He sent the proposal same day. Signed within the week.
Day 1 · 10:14am · Meridian call in 2 hours
David asked
"Tell me everything about my relationship with Meridian Healthcare."
Core · Entity page generated · 34s
142 email threads · 34 meetings · 12 OneDrive docs · 4 local files aggregated
Phase 1 Jan–Jun 2025 (complete) · Phase 2 started May 2026 (current)
CFO Patricia Novak: 67 emails · COO Robert Hsu: 48 emails · mostly strategic
Pipeline.xlsx: total invoiced $187,500 · Phase 2 margin 42% — second lowest active
Budget flag: Robert Hsu May 12 — "$120K vs $150K originally discussed"
Core — full client brief
Phase 1 closed June 2025 — Patricia Novak replied "excellent work." 6-month gap (subcontractor period). Phase 2 SOW signed May 2026, scoped to $120K (reduced from $150K) after Robert Hsu flagged a tighter board-approved budget.

Your 2026_Pipeline.xlsx shows 42% margin — second lowest active client. The Phase 2 Digital Transformation deliverable was last edited 2 weeks ago.
Evidence trail
Outlook · Patricia
Phase 1 feedback — "excellent work"
Jun 2025
Outlook · Robert
"tighter envelope — $120K vs $150K originally discussed"
12 May 2026
Excel · Pipeline
Meridian Phase 2 · $120K · 42% margin · "reduced from original scope"
Current
OneDrive
Meridian_DigitalTransformation_Framework.pptx — last edited 2 weeks ago
Active
Scenario C · Personal · Research

500 papers.
5 years. One question.

Lena Bergström — PhD researcher · Computational biology · Year 4 Writing her thesis literature review. She needs to trace the evolution of a technique across 5 years of reading. Normally a full day. She has 540 PDFs, 180 notes, emails with 420 collaborators, and years of lab meeting calendars.
Sources connected · 5 years of research
Gmail · 8,000 emails · 420 collaborators
Google Drive · 230 files · 3 research groups
Google Calendar · lab meetings · conferences
Local · 540 PDFs · 180 notes · Python code
What Core surfaced unprompted
None of the 23 papers address disordered regions. Core cross-referenced her January 2025 note flagging this as a future direction — then found Dr. Sofia Petrov, one hop from two of her collaborators, whose work directly fills the gap in her thesis. Lena had never heard of her.
Day 1 · Thesis writing · literature review chapter
Lena asked
"What papers in my collection use transformer architectures for protein structure prediction? Show me how the approaches evolved."
Core · 540 PDFs traversed · multi-hop graph query
Topic "transformer architecture" + "protein structure prediction" → 23 papers matched
Sorted chronologically · method evolution extracted from deep PDF summaries
Author edges: 4 papers from groups Lena has emailed directly
Drive: transformer_membrane_proteins_v4.tex — advisor Dr. Wei: "strongest chapter"
Gap identified: no papers address disordered regions · Jan 2025 note flags this
Core — literature brief
23 papers found. Foundational (2020–21): AlphaFold (connected to 8 of your papers) + ESM language model. Extension (2022–23): Baek et al. — your 2023 note labels this "the gap we should target." Current wave (2024–25): 4 papers from groups you've emailed.

Your draft extends the Baek approach to membrane proteins. Dr. Wei called it "the strongest chapter." Critical gap: disordered regions — zero coverage across all 23 papers.
Evidence trail · sources + 2-hop connections
PDF · Jumper et al.
AlphaFold — Nature 2021 · your most-cited ref · connected to 8 papers
2021
Notes · Jan 2025
research_ideas_2025.md — "disordered regions: future direction" · no follow-up
Jan 2025
Gmail · Dr. Wei
transformer_membrane chapter — "the strongest" · include HDBSCAN vs GMM
Mar 2026
Graph · 2-hop
Dr. Sofia Petrov · IDP + membrane proteins · co-author with Dr. Tanaka · never contacted
2024
PC
gave you computing
Internet
gave you connection
Mobile
gave you presence
Cloud
gave you scale
NeuralRecall
gives you continuity
For the first time, intelligence becomes persistent.
Core — My Digital Twin
NeuralRecall.ai · The fifth layer
Intelligence agents · Live

Three agents running.
7 seconds to extract.

Messaging agent · 13s
Entity agent · 25s
Action agent · 7s

Every piece of incoming information
automatically understood, tagged,
connected, and actionable.
10 / 16

Your intelligence compounds.

Month 1
Emails, calendar
documents
→
Month 6
Relationships, projects
patterns emerging
→
Month 12
Full intelligence
complete context
The longer you use it, the more irreplaceable it becomes.
11 / 16

The system learns
how you think.

Doctor
Diagnostic workflows.
Patient histories.
Clinical patterns.
Consultant
Client relationships.
Delivery frameworks.
Proposal logic.
Lawyer
Case precedents.
Negotiation history.
Clause libraries.
Investor
Deal flow signals.
Founder relationships.
Portfolio context.
Researcher
Literature networks.
Hypothesis trails.
Collaboration maps.
12 / 16
Detachable Data Agent

Your expertise.
Working while
you sleep.

A DDA is a structured slice of your intelligence — a living capsule of everything Core has learned about how you think in a specific domain. It can be attached to any AI session, shared with a collaborator, or listed on the DDA marketplace to earn passive income from your accumulated knowledge.
1
You use Core normally. Every query, every answer, every source connection teaches Core how you think in your domain.
2
Core structures your intelligence. After months of use, it identifies reproducible patterns — your frameworks, decisions, and domain expertise — and packages them as a DDA.
3
You list it. The DDA appears on the marketplace. Others — lawyers, consultants, researchers — query your expertise. You earn per query. You did nothing except use Core.
Live example · Month 14
Dr. Rami Khoury
Cardiologist · 14 years practice · Core user since Month 0
Core surfaced — unprompted
"Rami, I've identified 3 publishable intelligence assets from your clinical practice. Based on query patterns from similar professionals, these represent £1,800–£3,200/mo in passive yield. You wrote nothing. I structured it from 14 months of your sessions."
AF Management Protocol v2.3
340 clinical decisions · 14mo of rounds
1,247 queries
£1,240/mo
Post-TAVR 72hr Monitoring Framework
31% fewer readmissions vs standard protocol
891 queries
£680/mo
Cardiology Research Synthesis
156 papers · 4 clinical frameworks
612 queries
£420/mo
Month 1 total
£2,340/mo
Earning · Rami did nothing except use Core

Carry your expertise anywhere.

◉
Your Graph
Everything you know.
Structured. Private.
→
⬡
DDA Capsule
Session-scoped.
Ephemeral. Yours.
→
◈
AI Session
AI that knows you.
Revoke anytime.
Detach a slice of your graph. Attach it to any AI session.
And if you choose — mint it. License it. Let it work for you.
14 / 16
One category.
One winner.
Competitive positioning · 6 dimensions · Personal AI intelligence
ChatGPT
Resets every session		 Rewind AI
Records, doesn't understand		 Notion AI
Only what you wrote		 Mem.ai
Notes only, no graph		 Limitless
Audio only, no synthesis		 Core ✦
Persistent memory Remembers across sessions ✕ Partial Notes only Notes only Audio only ✓
Cross-source synthesis Email + calendar + docs + Slack ✕ ✕ ✕ ✕ ✕ ✓
Entity intelligence People, projects, orgs — structured ✕ ✕ ✕ Basic ✕ ✓
Runs locally / private Data never leaves your device ✕ ✕ ✕ ✕ ✕ ✓
Compounds over time Gets smarter the longer you use it ✕ ✕ Minimal Minimal ✕ ✓
Exportable intelligence DDA — your expertise earns for you ✕ ✕ ✕ ✕ ✕ ✓
Never explain
yourself again.
Core remembers.
Core connects.
Core · NeuralRecall.ai
23 / 23
Part II · The Product
Meet Core.
Core · NeuralRecall.ai · My Digital Twin
17 / 23
Core knowledge graph
The Knowledge Graph

Every signal.
Every source.
One living map.

Gmail · Slack · WhatsApp · LinkedIn
Google Drive · Spotify · Calendar
All connected. All queryable. All yours.
18 / 23
Core orbital twin view
Your Digital Twin

You, at the centre of everything you know.

19 / 23
Core answering under pressure
The Moment That Matters

20 minutes
to your biggest
meeting.

You asked
"What's my strongest angle based on our actual progress?"
Core answered
Your strongest angle isn't the tech — it's the velocity.
Core thought for 32s. It read your SHA draft, your WhatsApp thread, your Spotify session, and your investor deck — and connected what you couldn't.
20 / 23
Core evidence panel
How Core Thinks

Not hallucination.
Evidence.

Sources connected
Legalfab_SHA_Draft_v2 — deferred shares tied to $50M valuation
Gmail (Tuesday) — law firm pilot signal
Spotify history — 4hrs Deep Focus during deck edits
WhatsApp — 42 beta user messages, 48 hours
Every answer is traceable. Every insight has a source. Core shows its work.
21 / 23
Core entity clusters
Entity Intelligence

Core clusters
what belongs
together.

People. Messages. Events. Projects.
Core doesn't just store — it understands relationships.

Every cluster is a living model of how your world connects.
22 / 23
Core · Growing with you
42 goals.
106 entities.
182 agents.
Connected · Compounding · Yours
8/13
Sources connected
62%
Signals linked
∞
Memory
23 / 23
There is another version of you.
It remembers everything.
It sees every connection.
It never loses context.
It keeps learning.
It keeps growing.
One day...
it may know you better
than you know yourself.
Core.
My Digital Twin
NeuralRecall.ai
← → navigate