Yig — the headless agent system
Yig

For finance teams · Excel · Slack · Teams · CLI · API

You drive.
Yig drafts.

Close, forecast, and budget — half the manual hours, none of the workflow rewrite. Yig runs inside the Excel and Slack you already use. Every output passes through your team’s review. Audit trails stay clean.

Request access → How it works →
how-it-works.md — You drive. Yig drafts.
Step 1

Yig drafts.

IC matching, GL recon, variance bridges, schedule prep — drafted from the data already in your stack. No re-platforming.

Step 2

You review.

Every output flows through your team’s review. Comments, edits, accepts, rejections — all logged with reviewer and timestamp.

Step 3

You ship.

Approved outputs land in your close folder, board deck, or audit binder. Yig disappears. Your workflow stays yours.

HITL Human-in-the-loop, every step. No black box. No surprise automation.
pain-log.txt — Real user feedback
GL Reconciliation Yig Finance
"Every month-end, we manually reconcile SAP data against three Snowflake tables. That alone takes three days — before close can even start."
CFO Listed manufacturing group
Data Consistency Yig Finance
"We have four different Excel files tracking the same KPI. Every time the CFO asks which one is right, nobody in the room can give a straight answer."
Head of FP&A Internet company
Analyst Efficiency Yig Thinker
"My analysts spend 60% of their time cleaning and formatting data. Less than 40% on actual analysis. That is not what I hired them for."
VP Finance Multinational
time-targets.md — Indicative, scope-bound, target-not-achieved

Indicative time targets

Manual baseline vs. review-only with Yig. Targets, not achieved customer outcomes — see footnote.

Task Manual baseline With Yig (review-only)
Intercompany reconciliation¹ ~1.5 days ~2 hours review
GL reconciliation² ~6 hours / month ~45 min review
Variance bridge for board deck³ ~4 hours ~30 min review
Audit schedule prep⁴ ~2 days ~3 hours review

Targets reflect internal benchmarks and private-beta engagements. Actual impact varies by data quality, scope, and reviewer thresholds. Verified customer case studies will publish after the first paid Q-close ships end to end.

  • 1 Scope: 12-entity group, ~200 IC pairs, monthly close.
  • 2 Scope: 50-account TB, monthly close.
  • 3 Scope: P&L variance vs. budget, top-10 line items, board-pack format.
  • 4 Scope: lease, prepaid, accrual, and revenue schedules for a single annual audit.
Method Targets, scoped, footnoted. We will publish verified case studies as customers ship.
terminal.exe — Try Yig now
Demo mode · connect your data sources for live queries
demo@yig · v1.6.0
trust.txt — Architectural promises (provable today)

Your data, your stack

Run Yig locally or in your own cloud. No mandatory SaaS data plane. No vendor-side data lake.

Every step logged

Reviewer, timestamp, input, output, and diff written to an append-only audit log. SOX-friendly.

BYOK — bring your own model

Claude, GPT-class, or self-hosted. You own the inference contract. Switch models without losing workflows.

Architecture documented

Data flow, audit-log model, reviewer gate, and security boundaries — all published as public spec articles. Inspectable before purchase.
No vanity counters. No customer logos we did not earn. We list what is provable.
Setup Wizard — Join the Yig Waitlist

What is your biggest financial data challenge?

Select the one that best describes your situation — we'll recommend the right product.

connectors.exe — Vote for your stack

Roadmap vote · Yig Finance

Which data source does your team use?

Vote to determine which connector we build first.
Live results show the stack distribution of users on the waitlist.

SAP ECC / S4HANA Enterprise standard
47
Excel / Google Sheets Most common
44
Snowflake
38
Oracle NetSuite
24
Microsoft Dynamics
18
QuickBooks SMB
12
Google BigQuery
8
Power BI / Fabric Yigsemantic
15
One vote per user
206 engineers voted · live
Yig Finance · connectors roadmap
access.exe — Getting started

Closed beta · Python 3.11+

Once you have access

$ yig --version
yig 1.6.0

$ yig thinker init
   ●
  ● ●
→ thinker · 32 tools loaded
  model  claude-opus-4-7
  ctx    200k tokens

$ yig workflow gl_recon --period Q2-2026
∴ running gl_recon template…

Distributed via private beta · not on PyPI

Delivery surfaces

Runs where you work

# CLI / TUI
$ yig thinker run "Q2 close analysis"

# IM bots (Feishu / Teams / Google Chat)
@yig-thinker run gl_recon for Q2

# Excel / PowerPoint / Word
Yig Office Add-ins · BYOK · no backend

# REST API gateway
POST /v1/run · FastAPI · WebSocket
Products — Yig System Overview

Yig Thinker

Community and SMB flagship. Headless reasoning agent with 40+ tools, multi-step planning, and native IM integration. Runs in Slack, Teams, Feishu, and any terminal.

Learn more →

Yig Enterprise

Governance, SSO, audit trail, and multi-tenant management for institutions. Built for security teams and compliance-first organizations.

Learn more →

Yig Finance

Finance and data integration layer. Consensus across sources, structured extraction, and automated reconciliation. Sold separately.

Learn more →