For private equity operating partners

Find the EBITDA in the first 100 days — measured, not estimated.

The Portfolio Program measures how work is actually performed at up to three recently acquired companies at once. The OLi agent runs on each company’s desktops for four weeks; you get savable hours per company, converted to an EBITDA range, and a ranked list of what to do first.

$55,000 for up to 3 companies · 4 weeks of measurement · $17,500 per additional company

What you get

What each portfolio company’s report looks like.

Illustrative — synthetic data

Illustrative · synthetic data

Measured occupancy per agent, against an 80% target

40%team occupancy. No agent reaches 80%; 12 of 24 are below 40%.

Agent 11: 60% measured occupancyAgent 12: 55% measured occupancyAgent 10: 55% measured occupancyAgent 14: 53% measured occupancyAgent 18 (lead): 53% measured occupancyAgent 13: 52% measured occupancyAgent 22: 49% measured occupancyAgent 16: 49% measured occupancyAgent 09: 49% measured occupancyAgent 15: 47% measured occupancyAgent 24: 46% measured occupancyAgent 04: 44% measured occupancyAgent 23: 39% measured occupancyAgent 19: 37% measured occupancyAgent 06: 35% measured occupancyAgent 20: 34% measured occupancyAgent 01: 33% measured occupancyAgent 21: 32% measured occupancyAgent 02: 31% measured occupancyAgent 08: 28% measured occupancyAgent 05: 22% measured occupancyAgent 03: 20% measured occupancyAgent 07 (lead): 19% measured occupancyAgent 17: 17% measured occupancy
AgentTeam lead80% target
Run this for your team →

Illustrative · synthetic data

AI use cases and automations, ranked by measured hours

2,250–3,100 hsavable a month across eight use cases, from 21,584 measured hours.

  • Ticketing ↔ industry-specific platform lookups2,420 h · 485–725
  • Bank reconciliation1,450 h · 290–430
  • Ticket reply drafting1,380 h · 415–550
  • Call wrap-up & notes1,370 h · 340–480
  • General-ledger transaction entry1,090 h · 380–545
Measured hours / monthSavable (moderate)
68%of staff (108 of 160) pasted work into a public AI chatbot. Shadow AI, ungoverned.Run this for your team →

Illustrative · synthetic data

People online vs work arriving, by hour of day (ET)

5.5agents online in a typical hour; the work arriving keeps only 1.8 of them busy.

00:00 ET — 4.9 agents online, 0.7 busy01:00 ET — 5.0 agents online, 0.5 busy02:00 ET — 4.8 agents online, 0.4 busy03:00 ET — 4.9 agents online, 0.4 busy04:00 ET — 6.1 agents online, 0.4 busy05:00 ET — 6.1 agents online, 0.6 busy06:00 ET — 5.5 agents online, 0.8 busy07:00 ET — 5.6 agents online, 1.2 busy08:00 ET — 6.7 agents online, 1.7 busy09:00 ET — 6.2 agents online, 2.2 busy10:00 ET — 6.2 agents online, 2.7 busy11:00 ET — 6.8 agents online, 3.0 busy12:00 ET — 6.8 agents online, 2.9 busy13:00 ET — 5.6 agents online, 2.5 busy14:00 ET — 4.9 agents online, 2.0 busy15:00 ET — 5.0 agents online, 1.9 busy16:00 ET — 4.8 agents online, 1.9 busy17:00 ET — 4.4 agents online, 2.1 busy18:00 ET — 4.9 agents online, 2.8 busy19:00 ET — 5.5 agents online, 3.4 busy20:00 ET — 5.6 agents online, 3.7 busy21:00 ET — 6.1 agents online, 3.1 busy22:00 ET — 5.6 agents online, 2.1 busy23:00 ET — 4.9 agents online, 1.3 busy00061218
Agents onlineAgents busy with arriving work
Run this for your team →

Illustrative · synthetic data

Handle time: what agents log vs what OLi measured

7agents log under 3 min per case and 3 log over 15 min. Measured at the desktop: about 7.9 min.

Agent 22: logged 1 min, measured 8.8 minAgent 02: logged 0.9 min, measured 7.5 minAgent 24: logged 0.8 min, measured 7.2 minAgent 15: logged 0.9 min, measured 7.1 minAgent 10: logged 2.3 min, measured 8.3 minAgent 06: logged 2.2 min, measured 7.9 minAgent 20: logged 2.3 min, measured 7.5 minAgent 18: logged 7.9 min, measured 8.9 minAgent 07: logged 7.2 min, measured 8.1 minAgent 01: logged 8.1 min, measured 8.8 minAgent 03: logged 6.6 min, measured 7.3 minAgent 05: logged 6.8 min, measured 6.9 minAgent 19: logged 7 min, measured 7.1 minAgent 14: logged 9.2 min, measured 9.3 minAgent 21: logged 6.9 min, measured 6.8 minAgent 11: logged 8.5 min, measured 8 minAgent 16: logged 9.6 min, measured 8.9 minAgent 08: logged 9.1 min, measured 8.3 minAgent 12: logged 9.8 min, measured 9 minAgent 09: logged 7.9 min, measured 7 minAgent 23: logged 9.1 min, measured 8.1 minAgent 04: logged 17.8 min, measured 8.9 minAgent 13: logged 15.8 min, measured 6.7 minAgent 17: logged 17.7 min, measured 8 min

0–20 min per case, one column per agent

Measured at the desktopSelf-logged in the tracker
Run this for your team →

Invented organizations, synthetic numbers — the shape of the findings, not a real customer’s. Answer five questions and we run the same charts against your team’s size and mix. Run the simulation for your team →

The arithmetic

How measured hours become EBITDA — and enterprise value.

Savable hours are measured, not estimated. Capacity becomes savings only when it is taken — hires avoided, overtime cut, new volume absorbed without adding people — so the model carries that step explicitly.

  • Annual hours returnedsavable hours / month × 1227,000
  • Annual capacity value× loaded hourly cost$945,000
  • Realized EBITDA impact× 60% actually taken (hires avoided, overtime cut, volume absorbed)$567,000
  • Enterprise value at exit× 8× EBITDA multiple$4,536,000

Illustrative inputs — the default hours figure is the low end of our synthetic sample report. Replace them with each company’s measured savable hours and loaded cost; the portfolio roll-up does this for every company.

The offer

Portfolio Program

For private equity: a measured value-creation baseline across recently acquired companies, converted to an EBITDA range for each 100-day plan.

Scope
Up to 3 companies, 75 seats each
Measurement
4 weeks
Fixed fee
$55,000

100% of the assessment fee is credited toward your first-year OLi deployment and implementation if you sign within 60 days of the readout.

What you receive

  • A Workforce + AI & Automation report for each portfolio company
  • Savable hours by company, as a conservative–moderate range, converted to an annual EBITDA range from each company’s loaded labor cost
  • A ranked 100-day list per company: what to automate, re-staff, or govern first — and what each item needs from the company
  • Right-sized headcount and a re-staggered schedule where the work is queue-driven
  • Shadow-AI exposure across the portfolio: how much work is going into public AI tools, company by company
  • A portfolio roll-up and an executive readout for the operating partner and each company’s leadership

Inside the first 100 days

From approval to a ranked list in about six weeks.

  1. Weeks −2 to 0

    One security package, many approvers

    We send one security and data-handling package that every company’s IT team can reuse. Each company approves its own seats and gives employee notice; measurement starts with the first approvals rather than waiting for all.

  2. Weeks 1–4

    Measure every company in parallel

    The OLi agent measures the in-scope teams at each company continuously for four weeks — passive, across the whole team, anonymized at capture, isolated per company.

  3. Weeks 5–6

    Per-company readouts, then the roll-up

    Each company gets its report and readout. The operating partner gets the portfolio roll-up: savable hours, EBITDA range and the ranked 100-day list for each company.

  4. Day 60 onward

    Capture the capacity

    Automations, schedule changes and help at the point of work, delivered through OLi. The assessment fee is credited toward a deployment signed within 60 days of the readout.

Questions operating partners ask

Why not rely on management’s estimates or a consultant’s interviews?

Both rely on what people report, and interviews change how people work while they are in the room. The portfolio roll-up is only as good as the numbers under it, so Dataken measures how the work is actually performed — passively, at the desktop, for weeks, across the whole team — and the savable hours come with the measurement behind them.

Can we use this in diligence, before close?

Only with the target’s cooperation, because the agent is installed on the target’s desktops and its IT team has to approve it. In practice the Portfolio Program runs post-close, inside the first 100 days, when the numbers feed the value-creation plan directly.

What if one company has more than 75 seats in scope?

The program covers up to 75 seats per company. A larger company is priced as an enterprise program on its own and still rolls up into the same portfolio view.

Is this employee monitoring?

No. Activity is anonymized at capture and the findings describe the work — how long it takes, where it waits, what repeats — not a ranking of individuals. Each company gives its employees notice before measurement starts.

What does each company need to provide?

A named approver and IT contact, installation of the agent on the in-scope desktops, employee notice, and its fully loaded cost per FTE so savable hours can be converted into an EBITDA range.

How does the fee credit work across the portfolio?

The assessment fee is credited toward the first-year OLi deployment for any company that signs within 60 days of its readout.

Put measured numbers under the 100-day plan.

A 30-minute scoping call covers the companies, the teams in scope and the approvals.