Every portfolio company on the same axes — no matter how they report.
Twenty portfolio companies. Twenty different update formats. Some founders send clean monthly PDFs, some update a Notion page on a Tuesday whim, a few haven't reported since Q2. Before board prep you're digging through six months of emails trying to remember what that seed-stage founder's NRR looked like in March, and which Series A is bleeding headcount.
Your AI portfolio health agent reads every format each founder sends, normalizes to six standard metrics, and builds one unified dashboard. Red flags — burn spikes, growth deceleration, runway threshold breaches — hit your Slack the moment the founder's update lands. Board prep on any portco: one paragraph, five minutes, every number current.
You stop asking "how's the portfolio?" and start knowing.
Twenty portcos · twenty update formats · zero unified view.
Portfolio reporting is a patchwork. Your best-operating founder sends a monthly PDF with the same six metrics every time, clockwork. Two others live-update a Notion page whenever they remember. Three use Google Sheets with different column orders. One emails you a Loom on the 3rd of each month. A few haven't sent anything since their last round closed. The data is there — just scattered across twenty different artifacts that nobody has ever put next to each other.
That fragmentation is why your "portfolio review" every board cycle takes a full day. You open twenty different files, translate twenty different formats into your head, try to hold the comparison in working memory, and inevitably miss the seed-stage company whose burn doubled because you looked at their latest update two months ago. The normalization itself is the work — and nobody has time to do it manually every month.
Six metrics. Stage-aware benchmarks. Same axes, every company.
The agent reads whatever format each founder sends and translates it to six standard dimensions: revenue, burn, headcount, retention, cash position, growth rate. Each dimension is stage-aware — 15% MoM growth reads as excellent at seed and concerning at Series C. NRR of 112 is solid for B2B and irrelevant for B2C where cohort curves matter more. The agent applies the right benchmark to the right stage automatically.
Custom metrics — CAC payback, gross margin, the specific KPI that matters in your sector focus — can be added during deployment. The default six cover 90% of portfolio reviews. The other 10% is your edge.
One paragraph per company. What moved, what's on track, what to ask.
For each company, each month, the agent produces a one-paragraph synthesis: what moved this month, what's on track, what should be your follow-up question for the next board meeting. Not a data dump — a narrative shaped around the two or three things that actually changed. "Nova AI grew ARR 14% MoM with flat burn and no major hires — the sales efficiency improvement from Q1 is sticking. Ask about the enterprise pipeline conversion they mentioned last call."
Board prep on any portfolio company becomes a five-minute read. You walk into the board meeting knowing what moved, what to ask, and where to push — instead of spending the first 15 minutes of the meeting catching up on numbers the founder already shipped two weeks ago.
Updates process as they land. Red flags fire the same day. Monthly rollup on the 1st.
The agent processes each founder's update the moment it lands — PDF via email, Notion page revision, Sheet update, wherever it comes from. Company dashboards stay current to within 10 minutes of the founder's send. Red flags trigger the same day, not at month-end rollup, so you intervene while the problem is still fixable.
On the 1st of each month a full portfolio summary ships to your firm's Slack channel — every portco one paragraph, red flags surfaced at the top, non-reporters flagged with gap counts. Board prep becomes a 20-minute read, not a four-hour reassembly.
Three questions every GP raises first.
What if founders don't want us monitoring their data this closely?
The agent reads whatever founders already send — nothing more. It doesn't request new reporting formats, doesn't tighten cadence expectations, doesn't expose founders to additional scrutiny. Every founder currently sending you monthly updates will notice nothing different; the normalization happens downstream of what they already share.
Our stage benchmarks are firm-specific — how is this calibrated?
The default benchmarks come from broader Series-level datasets, but you can override per-stage or per-company during deployment. Most funds have stronger internal priors about what "normal burn" looks like for their seed investments than any external benchmark captures — the agent uses yours where you have them and defaults elsewhere.
What if we already have Affinity or Carta Portfolio Management?
The agent complements existing portfolio tools rather than replacing them. Affinity tracks relationships and deal flow; Carta tracks cap tables; neither does the one-paragraph monthly synthesis with stage-aware benchmarks and red-flag alerts. Many deployments run the agent alongside Affinity, with the agent reading Affinity's data as one of its sources.
AI portfolio health monitoring — answered.
How does the agent pull data from portfolio companies that report differently?+
The agent reads whatever format founders send: monthly email PDFs, Notion dashboards, shared Google Sheets, Figma boards, even Slack thread updates. It extracts the numbers from each format and maps them to your standard six-metric schema. Companies that don't send regular updates get flagged — you see that gap too, not just the companies who report well.
Which six metrics does the agent track by default?+
Revenue (ARR or MRR, depending on model), burn (gross and net), headcount, retention (NRR for B2B, cohort for B2C), cash position (months of runway at current burn), and growth rate (QoQ or MoM). These defaults cover 90% of what portfolio reviews need — you can add custom dimensions (CAC payback, gross margin, specific KPIs) during deployment.
What happens when a company doesn't report for a month?+
The agent flags non-reporters on the dashboard along with how many months they've missed. In most funds, silence is itself a signal — the companies that go quiet on updates are often the ones the partner should be calling directly. You see gaps, not just data.
How does the agent know what's "normal" for each stage?+
Stage-aware benchmarks come from two sources: your firm's own historical portfolio data (if it's in the agent's reach) and broader Series-level benchmark datasets we license. The agent compares a seed company's 15% MoM against seed benchmarks, not Series B ones. You can override the benchmark set per company during deployment if you have stronger priors.
Can board members or LPs see the same dashboard?+
You control distribution. The default is private to the partner team, but you can create filtered views for LPs (fund-level metrics only, no company-by-company) or board members (single-company view with just their company). Role-based access is configurable per user during deployment.
What counts as a red-flag alert?+
Four default triggers: burn spike of more than 30% vs 3-month baseline, growth deceleration of more than 40% vs 3-month baseline, NRR or cohort retention dropping more than 10 points, or runway falling below 6 months. All thresholds are tunable per company — a pre-revenue seed company has different normal than a Series C growth company.
How much does AI portfolio health monitoring cost?+
Included in every beeeowl deployment tier, starting at $2,000 for Hosted Setup. One-time payment — no per-portco fee, no per-partner charge, no monthly tier based on fund size. See the pricing page for the full breakdown.