Auquan is building AI for deep work in finance

FIRST PASS

Auquan raises $4.5M to help financial services firms automate complex knowledge work (BusinessWire)

Auquan announced a $4.5M addition to their seed round, bringing total funding to $8M. Their AI solution uses RAG and AI agent architecture to automate complex, knowledge-intensive workflows in finance — work that typically requires specialized expertise and sustained focus.

Key points:

1. Used by 25% of top 20 global asset managers, investment banks, and PE firms

2. Can generate a 30-page credit memo in 5 minutes vs. 5 days

3. Covers 550,000+ companies and 2M+ data sources across 65+ languages

4. Funding will go toward expanding engineering and sales teams

Merlin’s Notes: Although Auquan has been around for a while, the continued emergence of vertical-specific agentic platforms (Rogo, Mako, BlueFlame, Brightwave, Desia etc.) suggests that we’re experiencing a transition phase from AI 1.0 to 2.0, where we move from tackling routine work to true automation. This has several fascinating implications, but we’ll only highlight a few:

1. Market Efficiency Paradox: As more firms adopt these tools, we might see a "tech arms race" in finance. The initial efficiency advantage could ironically lead to more market inefficiency as firms desperately seek unique data sources and proprietary angles that their competitors' AI can't access.

2. Different Core Skills Needed: The most valuable junior employees might no longer be those who can grind through analysis, but those who can ask the right questions of AI systems, identify edge cases where AI might miss crucial context, and synthesize AI-generated insights into novel investment theses.

3. Deal Dynamics: The ability to analyze deals faster might crank up the pressure to make decisions faster, potentially leading to more competitive bidding processes and a greater emphasis on proprietary deal flow as speed advantages diminish.

So, the key question for financial services firms isn't ‘what tool should we adopt?’, but rather ‘how should we redesign our operating model around this new capability?’.

Only firms that rethink their entire approach will thrive in the very near future.

HIGHLIGHT

Assessing AI impact vs hype with John Stecher of Blackstone (a16z Fintech)

In an episode of "In the Vault," Blackstone CTO John Stecher discusses how AI is transforming private equity and investment management. Key highlights from his conversation with a16z's David Haber include:

• Blackstone's tech transformation: Technology has become the backbone of the firm, with AI and data infrastructure playing crucial roles in investment decisions and operations.

• AI deployment challenges: Implementing AI in financial institutions requires careful consideration of data security and permissions. Blackstone built custom infrastructure around Azure and AWS to handle these requirements.

• Model strategy: Rather than building proprietary LLMs, Blackstone focuses on creating security layers that allow them to switch between different models (GPT, Anthropic, etc.) based on specific use cases.

• Impact on analysts: While AI can now handle tasks like DCF models and document review, Stecher emphasizes that investment banking remains an apprenticeship business. AI serves as an enabler for analysts to work more efficiently rather than a replacement.

• Cross-asset class applications: AI is being leveraged differently across private equity, real estate, and credit investments. In credit, for example, it's used for parsing covenants and risk management, while real estate focuses on operational efficiencies.

Stecher sees AI as crucial for Blackstone's growth, enabling the firm to evaluate more investments and manage larger portfolios while maintaining quality.

However, he emphasizes that the core of the business remains its "insanely smart investment people."

JARGON BUSTER

USE CASE SPOTLIGHT

AllianceBernstein uses AI to help with everything

Since 2017, AllianceBernstein has been building a sophisticated AI and data science operation that's revolutionizing its investment processes.

Under the leadership of Andrew Chin, Head of Investment Solutions and Data Science, the firm has implemented AI solutions that are delivering measurable ROI and competitive advantages.

Key Applications:

1. Regulatory Filing Analysis

• AI processes 400+ company reports daily using Natural Language Processing (NLP)

• Automatically flags significant changes in strategy or management

• Creates investment signals to predict performance trends

• Real win: Caught a major retailer's strategy shift in their 10K filing before market impact

2. Investment Document Processing

• NLP summarizes 300-page offering memorandums

• Doubled analyst productivity in reviewing new opportunities

• Enables faster, more informed investment decisions

3. Compliance Automation

• AI assists with ERISA compliance checks

• Processes up to 50 new 150-page securities issues daily

• Improved operational efficiency by 50-75%

• Enhanced risk management through text-based validation

Impact:

• Saved hundreds of thousands of dollars

• Significantly improved analyst productivity

• Enhanced risk management capabilities

• Created competitive advantages in a $704B portfolio

The success of AB's AI integration shows how traditional asset managers can leverage technology to transform their operations while maintaining human oversight in critical decision-making processes.

As their head of investment solutions notes, this capability rebuild was essential to staying competitive in an evolving industry.

TOOL OF THE WEEK

Mako AI

Repeat after me: competition is good. Although we’ve already seen several AI platforms focused on finance (Rogo, BlueFlame, Desia, Reflexivity, Auquan, etc.), each player seems to be attacking the market slightly differently and it’s really cool to see.

This week, we came across Mako, a startup that’s billing itself as the "first AI Investment Associate". What makes Mako particularly interesting is its approach to being a true AI teammate rather than just another software tool.

Here are a few key features that caught our attention:

1. Processes internal firm data without requiring heavy implementation or cleaning

2. Pre-built connectors to common CRMs and file-sharing apps

3. Deploys in your own cloud environment with SOC 2 Type II Certification

4. Doesn’t use your data to train their models

5. Voice interface coming in 2025 (this is an exciting take and I think the timing could be great as we see more adoption of voice AI)