Cómo OffDeal usa Claude para cerrar 91 millones de dólares en operaciones de M&A con un equipo de cuatro banqueros

OffDeal es un banco de inversión nativo de IA que usa Claude para gestionar cada etapa del asesoramiento en fusiones y adquisiciones. Tras migrar al Claude Agent SDK e iterar en el diseño MCP, la empresa elevó la precisión de evaluación interna del 25% al 85% y consolidó 12 flujos de trabajo agénticos independientes en un único agente de propósito general llamado Archie. Con cuatro banqueros gestionando cada uno entre 5 y 8 operaciones simultáneas, OffDeal cerró 8 transacciones por un total de 91 millones de dólares en su primer año.

Impacto

85%

Precisión de evaluación interna tras migrar al Claude Agent SDK

12 → 1

Flujos de trabajo agénticos consolidados en un único agente

5-8

Operaciones simultáneas gestionadas por banquero

$91M

Valor de transacciones cerradas en el primer año

Up to 4 hours

Tiempo de ejecución autónoma del agente de búsqueda de compradores

Desafío

Doce flujos de trabajo agénticos independientes eran frágiles y tenían limitaciones de contexto: cada llamada a la API estaba aislada, las operaciones implicaban cientos de millones de tokens de contexto, y cualquier cambio en un procedimiento rompía el flujo correspondiente.

Solución

OffDeal migró al Claude Agent SDK y rediseñó en torno a principios MCP, consolidando 12 flujos de trabajo en un único agente de propósito general llamado Archie que mantiene el contexto completo de cada operación y se mueve fluidamente entre todas las etapas del proceso de asesoramiento.

Herramientas y tecnologías

Lo que dicen los líderes

Creo que somos el único banco de inversión del mundo que tiene más ingenieros que banqueros.

Ori Eldarov, CEO y Fundador, OffDeal
Entiende todo el contexto.

Regístrate para leer casos de estudio completos, acceder a métricas detalladas y recibir todos los reportes.

Historia completa

OffDeal is an AI-native investment bank built on the belief that the middle market — founder-led businesses with $1–20 million in annual profit — deserves the same quality of M&A advisory as large corporations, but has historically been too expensive to serve well. The firm pairs bankers with software engineers and uses AI to do the work that would otherwise require a large team.

"I think we’re the only investment bank in the world that has more engineers than bankers," said CEO Ori Eldarov. With five engineers and four bankers, OffDeal runs every stage of a deal through software: originating clients, sourcing buyers, building seller and buyer lists, writing confidential investment memoranda, and running diligence.

The early architecture was a collection of 12 independent agentic workflows, each designed for a specific task. Every API call was smart in isolation, but the system was brittle. When a standard operating procedure changed, the corresponding workflow broke. The deeper problem was context: a single deal can involve more than a hundred million tokens of information, and the models couldn’t hold enough of it to handle problems outside their predefined scope.

Migrating to the Claude Agent SDK and redesigning around MCP principles changed the architecture entirely. Twelve workflows collapsed into Archie, a single general-purpose agent that holds the full context of a deal and can move fluidly between tasks. Eval accuracy — measured against an internal benchmark — jumped from 25% to 85%. Failed API calls from context overflow dropped to zero overnight. The buyer sourcing agent now runs autonomously for up to four hours, researching potential acquirers across 10 different sourcing methods without human intervention.

In its first year, OffDeal closed 8 deals totalling $91 million in transaction value, with each banker managing 5–8 concurrent deals and meeting 2–3 new potential clients per day supported by AI-generated preparation materials. The firm is proving that a small, software-augmented team can operate at the output of a much larger bank.

Casos similares

A
Airtree
Reduced from 2 days to minutes
market & competitor research time

Airtree, a $2 billion Australian venture capital firm, deployed Claude Cowork as shared firm-wide infrastructure to unify fragmented data across tools like Notion, Slack, Google Drive, and Affinity. The team built custom Skills to automate board meeting prep, market research, and portfolio reporting — cutting multi-hour tasks down to minutes. What began as individual productivity gains quickly scaled into a collaborative system where Skills built by one person benefit the entire firm.

Financial ServicesWWebflowTTypeform
I
Intuit
Higher
helpfulness rating vs. non-claude experiences

Intuit integrated Claude via Amazon Bedrock into its Intuit Assist feature within TurboTax to generate plain-language explanations of tax calculations. The integration combines Claude's natural language capabilities with Intuit's proprietary tax knowledge engine, serving millions of customers during peak tax season. The result was higher helpfulness ratings and improved completion rates for federal tax filings.

Financial ServicesTechnologyIAIntuit AssistABAmazon Bedrock
B
Block
75% saving 8-10+ hours
engineers saving time weekly

Block, the financial technology company behind Square, Cash App, and Afterpay, deployed Claude as the default model in its open-source AI agent, codename goose, running through the Databricks Data Intelligence Platform. The system gives approximately 4,000 employees across 15 job profiles direct access to internal data, automated SQL generation, and AI-assisted code writing. Engineers report saving 8 to 10-plus hours per week, with goose adoption doubling within a single month.

Financial ServicesCGcodename gooseDDatabricks
IG
IG Group
70
hours saved weekly for analysts through ai-assisted query handling

IG Group, an international financial services firm, deployed Claude across teams to streamline content production, multilingual communications, and data analytics workflows. The rollout reduced reliance on external agencies, delivered a sub-three-month payback on the AI investment, and saved analysts 70 hours per week through AI-assisted query handling.

Financial ServicesCClaude
R
Ramp
1+ million lines
ai-suggested code implemented in 30 days

Ramp is a financial operations platform serving businesses with spend control, automated approval workflows, and expense processing. The company adopted Claude Code across its engineering organization, reaching 50% weekly active usage within months. In the first 30 days of broad adoption, engineers implemented more than one million lines of AI-suggested code while cutting incident investigation time by 80% through custom observability integrations.

Financial ServicesDDatadogSSnowflake
MF
Money Forward
80%
engineer adoption rate

Money Forward launched its MEPAR program to embed Claude Code across its engineering organization, achieving 80% engineer adoption with 70% using it daily. API endpoint implementation time fell from two days to five hours, and developer onboarding compressed from one week to one day. Early adopters reported saving approximately seven hours per week.

Financial ServicesTechnologyCCClaude Code
S
Stripe
1,370
engineers deployed

Stripe collaborated with Anthropic to create a signed enterprise binary of Claude Code, deploying it to 1,370 engineers with zero configuration. One team migrated 10,000 lines of Scala to Java in 4 days instead of 10 weeks.

Financial ServicesCCClaude Code
N
Nevis
5+
hours saved per advisor per week on administrative tasks

Nevis is an AI platform for wealth management that automates the administrative burden crushing independent financial advisors—from CRM data entry to meeting notes and client emails. Built on Claude Opus 4.5 and Claude Haiku 4.5, the platform serves firms collectively managing more than $50 billion in client assets and saves each advisor more than five hours per week on routine tasks. The result is a clear path for advisors to double their client capacity while maintaining service quality.

Financial ServicesNn8nCOClaude Opus 4.5