C
Crestline / ai
01 — Organisational System of Record

An AI that knows
your whole org.

Four layers of persistent memory — user, project, department, organization — synthesized in real time. Context never resets. It compounds forever.

LIVE · MVP-1· 4-LAYER MEMORY· 13 MODULES
CRESTLINE · CONSOLE
MAIN
// ⌘-K from anywhere · 19 tools loaded
📜 2 new in CONSTITUTION
· POLICY · Vendor approval policy → v4
$
● online· main
02 — Architecture

Built on four pillars

Every engineering decision must strengthen one or more. Each pillar is load-bearing — strip any one out and the system collapses.

PILLAR · 01
Verified Org Graph

Every person, dept, manager, project — typed FKs, audit trails, time-travel queries. Row-level access falls out of structure, never configured separately.

org
├─ dept-finance
│ └─ user-sarah → mgr:raj
└─ dept-eng
PILLAR · 02
Organisational Memory

Four Qdrant-backed layers — User · Project · Dept · Org. Forward-only, attributed at write time, retrieved by hierarchy-aware semantic + keyword search.

[USER]
[PROJECT]
[DEPT]
[ORG]
PILLAR · 03
Workflow Engine

NL-authored, org-graph-native automations. @manager_of(x), @dept-finance, @cxo resolved at runtime. n8n on three structural axes you can't out-build.

trigger ─→ query
└─→ email
└─→ wait
└─→ escalate
PILLAR · 04
Integration Mesh

One EventSourceAdapter interface — CSV, Postgres, Salesforce, REST live today. SAP, Workday, Oracle, Tally, Razorpay slated. The rest of the pipeline is source-agnostic.

SAP ─┐
WDAY ─┼─→ adapter
SFDC ─┤ │
REST ─┘ └─→ crestline
03 — Verified Org Graph

Build your org in motion

People, departments, managers, reporting lines — typed FKs with referential integrity. Tokens like @manager_of(Sarah) and @dept-finance resolve live against the graph.

INITIALISING
← swipe to explore →10 nodes
AV
Anika V.
Org Admin · L0
org
RM
Raj M.
Org Admin · L1
Finance
PN
Priya N.
Org Admin · L1
Engineering
DK
Dev K.
Org Admin · L1
People
A
Arjun
User · L2
Finance
M
Meera
User · L2
Finance
SK
Sarah K.
User · L2
Engineering
V
Vikram
User · L2
Engineering
N
Nisha
User · L2
People
K
Karan
User · L2
People
One typed FK away

Membership.managerId is the reporting-line foreign key. Cycle-safe writes, hierarchy-aware row-level access — permissions never configured separately.

Token vocabulary the AI speaks

@manager_of(x) · @dept-finance · @reports_of(x) · @cxo · @self. Same shape across workflows, messages, and the Constitution.

Time-travel built in

Every change to the graph audited. asOf=2025-12 returns the org as it was six months ago. Required for compliance and forensic audits.

04 — AI Hub

Memory that compounds

Every query fans out across all four memory layers in parallel. The Hub synthesises a single answer with inline citations to the exact memory items that grounded each claim.

AI HubREADY
What's slowing hire-to-onboard down this quarter?
C
ready · ask anything…
USER · layeridle
Your last review note
Onboarding playbook v3
Saved query: time-to-fill
PROJ · layeridle
Hiring sprint Q2
Background check vendor
Offer letter template
DEPT · layeridle
Talent ops weekly review
Recruiter dashboards
IT laptop SLA — 5d
ORG · layeridle
Hiring approval policy
DPDP consent log
Compensation bands
05 — Process Mining · MVP-1 Flagship

Find the $ leaking out of your org.

Upload an event log — or connect SAP, Salesforce, Tally. Crestline reconstructs the actual process, quantifies every bottleneck in real money, and names the people who can fix it. Grounded in your verified org graph — anti-hallucination by construction.

EVENT LOG · INGESTING
C-1742, intake, 2026-04-12 09:14, Anika V., ORACLEC-1742, kyc, 2026-04-12 09:42, Raj M., SAPC-1742, credit-check, 2026-04-14 11:08, Sarah, SAPC-1742, docs-req, 2026-04-14 14:22, Bhushan, SAPC-1742, docs-ver, 2026-04-17 16:01, Nisha, ORACLEC-1742, disburse, 2026-04-17 17:33, auto, FINACLEC-1743, intake, 2026-04-12 09:18, Anika V., ORACLEC-1743, kyc, 2026-04-12 10:11, Raj M., SAPC-1743, credit-check, 2026-04-15 09:45, Sarah, SAPC-1742, intake, 2026-04-12 09:14, Anika V., ORACLEC-1742, kyc, 2026-04-12 09:42, Raj M., SAPC-1742, credit-check, 2026-04-14 11:08, Sarah, SAPC-1742, docs-req, 2026-04-14 14:22, Bhushan, SAPC-1742, docs-ver, 2026-04-17 16:01, Nisha, ORACLEC-1742, disburse, 2026-04-17 17:33, auto, FINACLEC-1743, intake, 2026-04-12 09:18, Anika V., ORACLEC-1743, kyc, 2026-04-12 10:11, Raj M., SAPC-1743, credit-check, 2026-04-15 09:45, Sarah, SAPC-1742, intake, 2026-04-12 09:14, Anika V., ORACLEC-1742, kyc, 2026-04-12 09:42, Raj M., SAPC-1742, credit-check, 2026-04-14 11:08, Sarah, SAPC-1742, docs-req, 2026-04-14 14:22, Bhushan, SAPC-1742, docs-ver, 2026-04-17 16:01, Nisha, ORACLEC-1742, disburse, 2026-04-17 17:33, auto, FINACLEC-1743, intake, 2026-04-12 09:18, Anika V., ORACLEC-1743, kyc, 2026-04-12 10:11, Raj M., SAPC-1743, credit-check, 2026-04-15 09:45, Sarah, SAP
← swipe to explore →7 steps
Application Submitted
1,840 · 12m
ORACLE
Anika V.41%
Initial Validation
1,832 · 1.4h
SAP
Raj M.38%
$1.2M flows out
Credit Check
1,610 · 8h
SAP
Sarah K.67%
$8.2M flows out
Documents Requested
1,420 · 3.4d
SAP
Bhushan32%
$9.7M flows out
Credit Rejected
190 · 30m
ORACLE
auto88%
Documents Verified
1,205 · 1.2d
ORACLE
Nisha44%
$8.4M flows out
Disbursed
1,198 · 22m
FINACLE
auto92%
$8.4M flows out
Legend
Fast — under typical
Around typical
1.5–3× slower
Bottleneck (3×+ slower)
Dashed = system handoff
Bottleneck step
Has top performer
Sometimes skipped
Click any step to see who handles it · drag to rearrange
READY
06 — Workflow Engine

Workflows, but it knows your org.

Describe automations in plain English — Hindi works too. Five agents stream a validated spec. @manager_of(x), @dept-finance, @cxo resolve against the verified graph at runtime.

NL · /workflows/new
$
Intent
Planner
Drafter
Validator
Narrator
← swipe to explore →6 steps
trigger · schedule
FRI 18:00 IST
Asia/Kolkata
query · prisma
tasks { BLOCKED }
scoped to org graph
transform · group_by
@manager_of(assignee)
fan-out · ~12 groups
branch · age
task.ageDays > 7 ?
split · 2 paths
integration.email.send
manager digest
@manager_of(@self)
integration.email.send
CXO escalation
@cxo · subject prefix
READY
07 — Crestline Console

Command your org from one prompt.

Cmd-K opens the Console from anywhere. Slash commands run deterministic queries against the graph; free-form questions enter an agentic ReAct loop that plans, calls tools, and cites memory items inline.

CRESTLINE · CONSOLE
CONNECTED · MAIN
// session restored · 3 turns · 19 tools loaded
📜 2 new in CONSTITUTION since you last looked
· POLICY · Vendor approval policy → v4
· SOP · Credit-check SOP → v2
$
● online· cmd-k· /help· main
Cmd-K
Opens the Console from any dashboard route. Drag the top edge to resize.
15 slash commands
/whoami, /org, /team, /find, /tasks, /bottlenecks, /workflows, …
3 ACTION tools
create_task, dm, log_decision — gated by intent-word triggers.
08 — Projects · Module 3

Cockpits that design themselves.

Describe a tile in English — the AI Director emits a validated CustomTileSpec and the runtime renders it. Cockpits compose, remix, and persist. Not codegen — composable spec runtime.

READY
>
personaCFOFounder
Composable spec runtime
Discriminated unions for DataSource · Transform · Viz. Widen the union to extend — never freestyle code.
AI WBS generator
One-shot 4-level tree (Phase / Deliverable / Work Package / Task) streamed reading → outlining → detailing → validating → creating.
Smart suggestions panel
Proactively scans project state and offers 3-5 'you should add this' tiles with reason chips + one-click +.
09 — Project Research v2

Decision-grade candidate shortlists

Geography-aware query refinement → parallel Tavily fan-out → optional URL extract → structured ranked candidates with tradeoffs and a recommendation. Replaces 'one summary + action items' with something you can actually act on.

READY
>
Parallel sources
Justdial
queued
IndiaMart
queued
Sulekha
queued
Reviews · generic web
queued
tavilyExtract · top URLs
queued
collecting source signal…
10 — Data Insights

Drop the data, keep the meaning.

A multi-agent pipeline profiles every column, hypothesises, computes locally, visualises, and narrates. The LLM never sees rows in bulk — it picks the chart and writes the story. Server-side stats, AI only narrates.

READY
·
arr_by_segment_2024.csv · 5,142 rows
profile
hypothesise
compute
visualise
narrate
#1trendchart · bars

BFSI carries the org — but margin is in Pharma.

ARR by segment, last 12 months, ₹ Cr — only BFSI still double-digit.

actual projected
AI narrator
Awaiting compute…
What-if simulator
+0 sales hires
projected BFSI: 78.0 Cr (+0.0)
Ask the data
> Which segment will overtake Mfg first?
11 — Crestline Meet · Scribe

The AI that joins your meetings.

Crestline Meet runs in-app on LiveKit. Crestline Scribe joins as a real participant — listens, transcribes via Deepgram, and files every decision · task · blocker into the right place in real time. Closing the meeting publishes the notebook to org memory.

PN
Priya N.
SK
Sarah K.
RM
Raj M.
C
REC
Crestline· scribe
REC · 00:08:14
WAITING TO JOIN
CAPTURE QUEUE · LIVE
listening…
Routes
◆ decisionORG · decision register
▸ taskproject "Meeting captures"
⛔ blockerescalateBlockedTask → manager
12 — Crestline Constitution

The org's source of truth — hash-chained.

Policies, SOPs, MOMs, POs, Templates — authored by org or dept admins, read by everyone in scope. Every publish snapshots a hash-chained version; every body auto-embeds into the right memory layer.

LIBRARY · 289 DOCS
Library
search docs…
Org-wide·v4

Vendor approval policy

PUBLISHED92% ack

Vendor approvals > ₹40L require CFO + CTO sign-off. Concentration ceiling: no single vendor > 18% of OpEx without board note.

1. Scope. Applies to all procurement requests above the policy threshold issued by Crestline Org members from any department.

2. Approval cascade. Triggers route through @manager_of(@self) @dept-finance @cfo.

ack requireddpdp linked
VERSION CHAIN · SOC2-READY
publish.hook → embed body into crestline_org_memory
publish.hook → fire doc.published workflows
publish.hook → require ack from @dept-finance + @cfo
13 — Personal Cockpit · Mira

The AI that knows you — and your work.

Multi-agent productivity cockpit + trainer + analyst + skill-path designer. Five parallel agents stream chat, learning, preferences, artifacts, and inline cards. @me.signature, @me.po-template resolve across Messages, Console, and Workflows.

MIRA · 5 AGENTS · CHAT · LEARN · TUNE · ARTIFACT · CARDS
first-token < 500ms
Productivity heatmap · last 7d
Peak hour
10–12
Switches
7/d
Deep work
2.4 h
14 — Crestline · SOR (System of Records)

The constellation — 13 modules, one graph.

Each module is a verticalised intelligence layer that sits on top of your existing system of record (SAP / Workday / Salesforce / Tally). Independently sellable; the constellation is the moat.

MOD · 01
Core-Org
CEO · COO
LIVE
MOD · 13
Process Mining
CFO · COO
SAP Signavio · CelonisLIVE · MVP-1
MOD · 02
People
CHRO
SAP SuccessFactorsQ4 2026
MOD · 03
Projects
COO · PMO
SAP PPMQ1 2027
MOD · 04
Finance
CFO · FP&A
SAP FICO (intel layer)Q1 2027
MOD · 05
Sales
CRO · VP Sales
Salesforce (intel)Q2 2027
MOD · 06
Operations
COO · Procurement
SAP AribaQ2 2027
MOD · 07
Compliance
CCO · Risk
SAP GRCQ1 2027
MOD · 08
Customer
VP CS
SAP Service CloudQ2 2027
MOD · 09
Engineering
CTO · VP Eng
(white space)Q3 2027
MOD · 10
Knowledge
CIO
GleanQ3 2027
MOD · 11
Strategy
CEO · Board
(white space)Q3 2027

one verified org graph · one memory mesh · one workflow engine · one integration mesh — 13 modules consuming all four

16 — Make the move

Ready to make your org legible?

30-minute walkthrough with a founder. We bring real data — yours or ours. You walk out with one named bottleneck and a ₹ number against it.

crestlineintelligence · proprietary · india