AI Partner for Investment Firms

    Your team should be using AI on every deal. We make that happen in week one.

    We configure Claude and ChatGPT with tailored system prompts, PE-specific agent skills, and role-based playbooks for your deal team, IR, and finance. No new software. No six-month rollout. Working drafts on real deliverables by Friday.

    Citations built into every template. Nothing goes external without review. Your data never trains the model.

    Book a 30-Minute Workflow AuditWhat you get in week 1

    What you get in week one.

    Most teams see value in the first few days, because we start with real workflows and real deliverables, not roadmaps.

    • ✓4-6 role-based AI playbooks with tailored system prompts installed inside Claude or ChatGPT Projects — each pre-loaded with your firm's context, PE terminology, and formatting standards
    • ✓Pre-configured agent skills for core PE workflows: deal screening, portfolio monitoring, memo drafting, and LP communications — ready to use on day one
    • ✓Copy/paste templates and examples built around your team's actual deliverables
    • ✓Firm "rules of the road": what to use AI for, what not to paste, how to review outputs
    • ✓Live training session plus recurring office hours
    • ✓Shared output folder with naming and version conventions (Working Draft vs. Approved)
    • ✓Citation fields built into every template where facts, figures, or claims appear

    Built for firms that handle confidential data.

    Your playbooks aren't just prompts. They're a system designed around your NDAs, your compliance requirements, and your team's trust.

    🔒
    Zero-training guarantee
    Claude and ChatGPT Team/Enterprise plans never use your inputs to train models. Your deal data stays yours. We put this in writing.
    👥
    Human-in-the-loop, always
    AI never sends anything. It only drafts. Every output is reviewed by a named team member before it goes anywhere. No exceptions.
    📑
    Citation fields on every claim
    Facts, figures, and quantitative claims carry source references by default. Your team knows where every number came from.
    📚
    Rules of the road
    Clear, firm-specific guidelines on what to paste, what not to paste, and how to handle sensitive deal materials. Reviewed in training on day one.

    Use it. Share it. Trust it.

    Every workflow follows the same path. Start with personal value, scale what works across the team, then systematize the outputs you trust enough to send externally.

    Use it.
    Start

    Personal AI playbooks for your team. Working drafts on real deliverables in week one. Your team gets comfortable, sees immediate value, and builds the habit.

    What this looks like
    Analyst runs a CIM through the Deal Brief playbook before reading it manually
    IR drafts LP update blocks in 20 minutes instead of two hours
    Partner gets a one-page meeting prep before every call
    What this looks like
    Deal team shares one CIM extraction format across all active opportunities
    Monitoring alerts go to the whole team, not just one person's inbox
    New hires get the same playbooks on day one
    Share it.
    Scale

    The playbooks that stick become shared team workflows. Consistent inputs, cited outputs, same process for everyone. No more "my version vs. your version."

    Trust it.
    Systematize

    When outputs go to IC, LPs, or external parties, we add governance: named reviewer, version history, and a single source of truth. Safe to send outside the firm.

    What this looks like
    DDQ responses drafted from canonical, approved answers
    LP letters generated from monitored portfolio data with audit trail
    IC memos carry version history and named reviewer sign-off

    As AI capabilities advance, so do your playbooks. What starts as a draft assistant today becomes an automated workflow tomorrow. We manage that evolution so you don't have to.

    Does this require new software?
    Use itNo. Playbooks and training inside Claude or ChatGPT. Nothing to install.
    Share itSometimes. Lightweight internal tooling for the workflows that stick.
    Trust itYes, when you need approvals, versioning, and a system of record.

    Example playbooks by team.

    Each playbook runs on a tailored system prompt built for your firm — encoding PE context, your formatting standards, and compliance guardrails. These accelerate first drafts. Your team stays in control.

    Investment
    CIM to Deal Brief
    CIM PDF → 1-2 page draft
    Executive summary, key financials, red flags, and initial thesis fit questions. Page-cited.
    IC Prep Pack
    Deal docs → structured brief
    What we know, what we don't, recommended next diligence steps. Ready for committee.
    Call Notes to Memo
    Raw notes → memo skeleton
    Structured memo with headers, supporting quotes, and flagged gaps in the thesis.
    Meeting Prep
    Company name → 1-page brief
    Thesis alignment, key questions, portfolio overlap, recent news. Ready before every call.
    Finance / Accounting
    KPI Variance Commentary
    Monthly numbers + notes → narrative draft
    Turns raw financials into commentary paragraphs for internal and LP consumption.
    Portfolio Summary Roll-up
    Portco updates → consolidated view
    Combines updates across companies into a single narrative with consistent formatting.
    Investor Relations
    LP Update Blocks
    Portfolio themes → quarterly language
    Consistent, reusable language per portfolio theme. Eliminates starting from scratch each quarter.
    News to "So What?"
    News link → brief + implications
    What happened, why it matters for our portfolio, suggested follow-ups. Ready to share internally.
    DDQ Draft Assist
    DDQ questions → drafted responses
    Pulls from past responses and firm materials. Flags gaps and inconsistencies for review.
    These are starting points. After week 1, we iterate based on what your team actually uses and what they ask for next.

    How we make adoption stick.

    Trigger-mapped playbooks

    Each playbook is paired with a specific moment in your week: CIM arrives, deal meeting tomorrow, LP letter due. The workflow is the reminder.

    Training plus repeat office hours

    One session to launch, then monthly office hours so questions get answered and new playbooks get added organically.

    Templates that enforce structure

    Citation fields, output labels, and formatting conventions are embedded in every playbook. Good habits by default.

    Weekly iteration on what sticks

    We track which playbooks get used and which don't. Kill what doesn't work, double down on what does.

    PE-specific logic, not a redundant subscription.

    Enterprise AI platforms charge for software your team may never fully adopt. We use the tools you already pay for and focus the investment on what actually matters: playbooks built for your workflows, training that sticks, and a partner who stays on to iterate.

    You're paying for PE-specific logic and ongoing advisory, not a platform license.

    Typical engagement: $20K to get started (4-6 playbooks + training + rules of the road), then $5K/month for ongoing advisory, office hours, and new playbook development. Most firms invest $80-120K in year one and save multiples of that in analyst hours.
    PlatformAnnual
    Hebbia$50-80K+5-10 seats
    Metal$50-100K+per fund
    Blueflame$50-100K+enterprise
    Welcome.AI$20K + $5K/moall playbooks

    Common questions.

    Do we need enterprise licenses to start?
    Not necessarily. We can start with Team-tier seats on Claude or ChatGPT. If adoption sticks or you need SSO and admin controls, we'll recommend an enterprise upgrade at that point. Most firms start with 5-10 seats.
    What about our NDAs and confidential deal data?
    Claude and ChatGPT Team/Enterprise plans contractually guarantee your inputs are never used for model training. We build firm-specific rules of the road on day one: what to paste, what not to paste, and how to handle sensitive materials. We only move to shared or governed workflows when the firm is ready.
    Does this replace our analysts?
    No. This accelerates drafts and analysis. Your team still reviews, edits, and makes every decision. Think of it as giving each person a research assistant that works instantly, not a replacement for judgment.
    What happens after week 1?
    We stay on as your AI partner through a monthly retainer: office hours, playbook updates, new workflow identification, and usage coaching. When a playbook proves valuable enough to share or systematize, we scope that as the next step.
    How many workflows do you set up?
    We scope 4-6 playbooks for week one, focused on the highest-frequency, highest-pain tasks we identify in discovery. Additional playbooks get added monthly through the retainer based on what your team asks for and what the usage data shows.
    Why is this so much cheaper than Hebbia or Metal?
    We don't build proprietary software on top of your data. We configure the enterprise AI tools you already pay for — tailored system prompts, agent skills, PE-specific playbooks — and add training and ongoing advisory. You're paying for our expertise and your team's productivity, not a redundant platform license.

    Start with a conversation.

    Thirty minutes. We'll tell you where AI fits, where it doesn't, and whether we're the right partner.

    Book a Workflow Audit
    Or email directly: [email protected]