For institutional allocators

Diligence and reporting at the speed of capital.

OCIOs are running 50+ manager reviews per cycle on spreadsheets and PDF readers. Aktus deploys an analyst that reads PPMs, normalizes performance, and writes IC memos against your house view, with citations.

10 days
From scoped pilot to first full IC memo
Manager reviews per analyst, per cycle
Compliance-ready
Memo trail with full citation chain
50+ GPs
Normalized in a single deployment
/ 01
Who we serve

Built for the teams deploying capital on a deadline.

Aktus deploys inside the allocators whose diligence packages, IC memos, and performance reporting are the operational fabric of institutional alts allocation.

OCIOs

Outsourced CIOs

Multi-billion-dollar OCIOs running endowment, foundation, and family-office mandates.

  • Manager diligence at scale
  • House-view normalization
  • IC-ready memo production
LPs

Sophisticated direct LPs

Pension and sovereign teams allocating direct to GPs, increasingly without consultants.

  • Multi-cycle pacing models
  • Performance attribution
  • LP-side fee analysis
FoFs

Funds of funds

Boutique FoFs differentiating on speed and thesis, against a wall of incoming PPMs.

  • PPM screening pipeline
  • Track record normalization
  • GP-by-GP underwriting
/ 02
What Aktus automates

From PPMs to a memo your IC reads.

A bespoke AI worker reads the same diligence stack your senior analyst reads, and produces the memo your IC would read, with every claim cited to a manager document.

Inputs the worker ingests
  • Private placement memorandums and DDQs
  • Manager presentations and data rooms
  • Historical track records and benchmarks
  • Reference call notes and ODD findings
  • Your house view, target return, and pacing model
  • Prior IC memos and post-mortems
Outputs the worker produces
  • IC-ready memos with citation chains
  • Performance normalization across vehicles
  • Fee load analysis vs. peer set
  • Risk factor extraction and side-letter map
  • Reference call discussion guides
  • Pacing-fit and portfolio overlap assessments
What stays with your team: conviction, the reference calls, and the allocation decision. The AI worker removes 70% of manual document parsing and normalization.
/ 03
How it works

Five steps from your house view to a live IC memo.

Deployment maps onto your allocation calendar. We start with one strategy, calibrate against three prior memos, and ramp into your next IC cycle.

01

Scope

We pick one asset class and one upcoming allocation. You name the diligence depth your IC requires.

Scope exampleDirect lending · Q2 IC cycle · 3 finalists.
02

Ingest

Data rooms land in a sovereign environment. Our AI worker builds a typed graph of every manager, vehicle, and track record.

ExamplePPMs, DDQs, performance, side letters, ref calls.
03

Calibrate

We reproduce three prior IC memos against the worker. Variance report surfaces every disagreement with your senior analyst.

ExampleWorker reproduces 92% of senior-analyst conclusions.
04

Deploy

Our AI worker runs the live cycle. Memos land in your IC packet. Your CIO signs off in their normal workflow.

ExampleLive for Q2 cycle on direct lending strategy.
05

Expand

Additional strategies added by configuration. Worker generalizes across asset classes, vehicles, and house-view evolution.

ExamplePE buyout & infra added by Q3.
You sign off on the worker before a single memo reaches the IC. Every output is reviewable; nothing is a black box.
/ 04
Why this needs Aktus

Frontier models can’t do institutional diligence alone.

Aktus makes them production-ready for Alternative Assets.

The frontier model problemManager diligence is a long-horizon, citation-heavy, house-view-anchored reasoning problem. A general model has no view; an OCIO’s value is the view. Aktus is the difference between a model that summarizes a PPM and a worker that underwrites it.
What Aktus solves

House-view anchoring

The worker reasons against your target return, your pacing model, and your conviction framework, not a generic benchmark.

Your view is a versioned, evaluable artifact.
Where models alone fail

Track-record normalization

Cross-vintage, cross-vehicle performance comparison requires structured work models can’t do reliably from prose.

Aktus uses typed extractors and audited math.
Worse than humans

Reference-call synthesis

A general model can summarize a call but can’t cross-check it against the PPM, the DDQ, and the prior cycle.

Aktus runs structured cross-document checks.
An AI diligence worker is not a model. It is a model wrapped in your house view, governed by an evaluation harness, and run inside your perimeter.
/ 05
Context orchestration

How our worker stays accurate.

Aktus runs every worker against a continuously evaluated context graph, your house view, your prior memos, your reference network, so the worker’s answers track your firm’s conviction, not the public internet.

01

Manager graph

Every PPM, DDQ, and track record is parsed into a typed graph with version history and vehicle-level provenance.

02

House view

Your target return, pacing model, and conviction framework are first-class objects the worker queries.

03

Eval harness

A regression suite of prior memos runs on every model change. Deployment is gated on consistency with your senior analyst.

04

CIO review

Memo-level outputs route to your CIO. Their edits update the house view for the next cycle.

/ 06
Economics

More managers reviewed. Same analyst team.

OCIO economics are bounded by the number of managers each analyst can underwrite in a cycle. Aktus changes the ratio.

Diligence today
  • ×Senior analyst time consumed by PPM reading
  • ×Manager throughput capped at 12–15 per cycle
  • ×Performance normalization done in Excel, manually
  • ×New asset classes require new analyst hires
  • ×IC memo quality varies analyst-to-analyst
With Aktus inside
  • Senior analyst time spent on conviction and refs
  • Manager throughput 4× without analyst hiring
  • Performance normalized to your house benchmark
  • New asset classes are a configuration, not a hire
  • IC memo quality reflects the firm, not the analyst
Managers underwritten per analyst, per cycle
60%
Reduction in time from data room to IC memo
< 90 days
Payback on a typical OCIO deployment
/ 07
Sovereign deployment

Your perimeter. Your data. Your fiduciary duty.

Aktus runs inside your environment, against your documents, under your audit trail. NDA boundaries are not crossed; primary documents never leave your boundary.

Sovereign by default

Deployed in your VPC or on-prem. Aktus operators see logs and metrics, never PPMs or DDQ responses.

  • SOC 2 Type II
  • AWS, Azure, GCP, on-prem
  • Customer-managed keys

Fiduciary audit trail

Every memo carries hash-chained provenance, committee can trace any claim back to the manager document it came from.

  • Immutable provenance log
  • CIO sign-off captured
  • Exportable for compliance review

Model-agnostic

Run on frontier APIs, your private inference, or hybrid. Switch underlying models without redeploying the worker.

  • Anthropic, OpenAI, on-prem
  • Per-task model routing
  • Versioned, evaluated upgrades
Compatible with your alts stack
Backstop
Dynamo
Solovis
Albourne
Aladdin
Cambridge

Allocate at institutional speed.

Your investment expertise. Our AI operations engine. Scale your OCIO without scaling your back office.