Run your fund on your own data.
Open-source portfolio reporting for venture funds, syndicates, and angels. Forward the founder updates you already get, and AI turns them into portfolio dashboards and LP reports. You self-host it, so your data stays yours and your AI keys stay yours.
How it works
Forward the updates you already get
Give founders an inbound address and send investor updates to it in any format, PDF, Excel, PowerPoint, or plain text. AI identifies the company, extracts the metrics, and flags anything it's unsure about for you to confirm.
Import the rest of your portfolio
Paste text, upload files, or bulk-import a spreadsheet. The same pipeline handles metrics, investments, and cash flows, so a quarter of history lands in minutes instead of an afternoon of copy-paste.
Watch the dashboards build themselves
MRR, burn, headcount, and your custom KPIs chart over time from the updates as they arrive. One source of truth for company health, with the numbers your admin never touches sitting right next to fund performance.
Report to your LPs from live data
Draft quarterly letters, print investor report cards, or freeze a snapshot for the archive. TVPI, DPI, and Net IRR compute from the investments you've recorded, so an LP report is a click, not a rebuild.
Who it's for
The same platform, framed for how you actually run. Start where you are.
SPVs
Run a single-deal vehicle like a fund: track the one company, its LP capital, and K-1s.
Solo GPs and emerging managers
The back office you don't have. Report to LPs from live data, with real books when you want them.
Angels and syndicates
A wide portfolio of small checks, tracked from the founder updates you already forward.
Micro and seed funds
The full stack, screening through LP reporting and real books, on infrastructure you own.
Turn on what you need
Monitoring is the core. Around it sit the other workflows a fund actually runs, each a switch that's off by default. Run the piece you need this quarter and ignore the rest until you don't.
Portfolio monitoring
The core, and the part most funds start with. Automated metric extraction from founder updates, a review queue for anything uncertain, and a real-time dashboard of company health.
Inbound deal screening
Cold pitches and partner intros sent to your inbound address get classified, scored against your thesis, and queued in a deals pipeline. Add a public submission form if you want founders to come to you.
Diligence and memos
An AI agent reads the data room, runs outside research, puts open questions to your partners, and drafts a structured memo with provenance on every paragraph, rendered to Word or Google Docs. The rubric and memo structure are yours to edit per fund.
LP capital tracking
Paste a statement with commitments, called, distributions, and NAV, and AI maps the columns into a dated position. Each import is stamped with its as-of date, so you can pull a capital account as of any date you like.
Fund accounting
When you want LP numbers coming from real books, turn on a double-entry ledger per vehicle. Import a bank feed, book capital calls and distributions, and close a period to allocate income into each partner's capital account, accruing note interest and carry along the way.
Compliance calendar
A regulatory calendar tuned to your fund's registration and structure. Answer a short questionnaire and it tracks the filings that apply to you, SEC forms, tax deadlines, Form D, and AML, so a deadline lands on a calendar instead of arriving as a surprise.
LP portal
A private, fund-branded login where your LPs read and download capital account statements, quarterly letters, and fund documents. Advisors and accountants you've authorized come along automatically, and an AI analyst answers questions using only that LP's own materials.
Lightweight CRM
Track intros, strategy, and qualitative work alongside the numbers. BCC your inbound address on a conversation and the interaction logs itself.
Model what this doesn't
This tracks and reports on live data, and it stops there on purpose. When you need to model a cap table, an exit, or fund returns forward, that's what the Hemrock models are for. Same builder, and built to edit with AI.
Cap Table and Exit Waterfall
Per-company ownership and dilution across rounds, and what each holder takes home at any exit.
Fund Economics Tool
Management fees, carry, and net-to-LP returns, modeled before you report them.
Venture Capital Model
A full fund forecast, deployment pace through exits and returns.
Venture Valuation Tool
Value a company before you write the check.
An AI analyst, on your AI
Every part of the platform has an AI analyst behind it, and it runs on your AI, not mine. Bring your own key for Anthropic, OpenAI, Gemini, or OpenRouter, or point it at a local model through Ollama, and switch providers whenever you want. Your fund data goes to the provider you chose and nowhere else.
Ask it anything about a single company or the whole portfolio, then put it to work. It drafts LP update letters from your metrics and notes, categorizes bank transactions into journal entries for the ledger, proposes valuation marks, and answers your LPs' questions in the portal from that LP's own documents.
It works inside your access controls, never around them. The analyst only sees and acts on what the person using it is allowed to see. An LP's portal analyst answers from that LP's materials alone, advisors and accountants see only the accounts they're authorized on, and anything the AI writes to your books is a draft you approve before it posts.
Connect it to your stack
Your reporting data doesn't sit in a silo. Pull it into the tools your fund already runs, and let your AI reach it directly.
API
NewA fund API key drives the full tool surface from your own scripts and headless clients: companies, metrics, investments, LP positions, fund performance, and the ledger, scoped to what the key can reach.
MCP server
NewThe same surface over MCP, so Claude and other AI clients query your portfolio directly. Bring your data to your own AI instead of handing it to someone else's.
Webhooks
Coming soonSubscribe to events and push them into your own systems the moment they happen.
Affinity
PrebuiltA prebuilt integration that pulls deal and relationship data from Affinity into your portfolio and diligence views. One-way for now, Affinity into Hemrock.
Why this exists
Most portfolio reporting platforms lock your data in their database, run it through their AI, and charge per seat so half your team can't log in. Fund managers shouldn't have to choose between good tooling and owning their data.
This is the whole reporting platform, deployed on your own infrastructure: your database, your AI keys, your domain. No per-seat fees, no black-box AI training on your portfolio, no vendor lock-in. It's a real double-entry ledger when you want one, not a read-only dashboard bolted on top of someone else's numbers, and it exposes an MCP endpoint so the AI of your choice can query the portfolio directly. Ask it anything, about benchmarks, trends, or a single company.
I built it out of my own fund reporting work and made it something any manager can self-host. It keeps growing because I keep running into the next thing a fund needs.
How this is different
Own your data
Self-hosted on your infrastructure, your database, your domain. Nothing lives in someone else's multi-tenant cloud.
Bring your own AI
Your keys and your choice of Anthropic, OpenAI, Gemini, or a local model, with an MCP endpoint so your AI queries the portfolio directly.
Real books, not just a dashboard
A double-entry ledger per vehicle with period close and financial statements, so LP numbers come from books rather than a spreadsheet on top.
No per-seat fees
Your whole team and every LP logs in. Pricing doesn't punish you for adding people.
Make it yours
There are two ways to shape it around your fund. From settings, adjust colors, labels, and styling with no code, so it looks like your firm out of the box.
And because the whole thing is open source, you can go further than any settings panel. Install it, open the source, and change any part of the app itself, the look, the workflows, the functionality, whatever your fund needs that the defaults don't cover. Most platforms let you upload a logo. Here you own the code and edit it like you'd edit one of my financial models, with your AI doing most of the work.
Pricing
The software is free and open source under the Apache License 2.0. Run it yourself, or have me set it up and run it for you.
Self-hosted
Free
Run it on your own accounts
- Free and open source (Apache 2.0)
- All your funds, SPVs, and team, no per-seat fees
- Your infrastructure, your domain
- Roughly $0 to $75 a month in running costs
Setup and support
From $2,500
One-time, I deploy it for you
- Deployed on your infrastructure and accounts
- Setup, onboarding, and data migration included
- Ongoing support available
Hosted
From $250/mo
I host and maintain it, cancel anytime
- No infrastructure for you to run
- Onboarding your data and process included
- Ongoing support included
Common questions
Do I have to use all of it?
No, and most people shouldn't. Deal screening, LP tracking, the LP portal, and fund accounting are independent switches, each off by default. Track LP capital from pasted statements with no accounting behind it, keep full books only for the vehicles you run yourself, or land anywhere in between. Portfolio monitoring works on its own.
Doesn't my fund administrator already give me portfolio reports?
A fund admin keeps the books of the fund. The core of this is the layer underneath that, the operating data of the companies you're invested in, which is the part most admins never touch. If you do want the accounting too, the ledger runs alongside your admin, or in place of one for the vehicles you handle yourself.
Is self-hosting a lot to maintain?
Less than it sounds. It's a standard web app you deploy once, and the quick start walks through it in an hour or two. If you'd rather not touch infrastructure, I can deploy it on your accounts, or you can go on the hosted subscription.
Is my portfolio data private?
Yes. You deploy on your own infrastructure with your own database and your own AI keys. Nothing is shared with other funds or used to train anyone's model. You own and control everything.
Which AI models does it use?
You bring your own key and pick the provider: Anthropic, OpenAI, Google Gemini, or OpenRouter, or a local model through Ollama. Switch whenever you want. AI powers the metric extraction, summaries, diligence, and letter drafting.
Can I modify the code?
Yes. It's open source under the Apache License 2.0. Use it, change it, and deploy it freely, for your own fund or commercially. I'm around to help with modifications if you want them.
