The perspective
Porter’s Five Forces has helped investors assess what protects a business’s ability to compete profitably, including the barriers faced by new entrants. Applied AI can lower some of those barriers by making parts of a service easier to reproduce. For us, that sharpens the investment question: which strengths will remain difficult to replicate, and how can the business build on them to deliver more value to customers?
At Astira, we back specialist services businesses whose customers rely on their expertise, information and judgment. We partner with founders and management teams to build on those strengths and support the next stage of growth.
Applied AI can help those businesses serve more customers, improve delivery and develop their offering. We assess both that opportunity and the pressure it may put on their existing advantages.
Those judgments shape which businesses we back and the priorities we agree on with management during ownership.
How applied AI changes the investment case
Applied AI can lower barriers to entry by reducing the time and resources a new competitor needs to deliver routine work. Earning customer trust, handling difficult exceptions, and delivering reliably can remain harder and will provide some insulation to incumbents who have built this expertise over time. We assess which barriers will still matter as the tools become more widely available.
Buyer bargaining power can also increase where applied AI makes competing services more comparable and customers have credible alternatives. They may expect lower fees or more service for the same price. A productivity gain does not automatically become a stronger margin. We ask what enables a business to retain a share of the value it creates.
We consider opportunities in two broad buckets:
- Redesigning existing work. How could the business use applied AI to improve the quality, responsiveness, or economics of the service it already delivers?
- Developing additional offerings. Where could its expertise and customer relationships support a new service that addresses unmet customer needs?
In each case, we assess what gives the business an advantage: the specialist knowledge, information and operating capabilities to redesign delivery, and the customer trust to win additional work.
General-purpose tools can support either opportunity. The investment decision depends on what the business can deliver, the resources required, and the risks involved.
We consider those questions before investing and develop the priorities with management during ownership. The ambition is to make the service more valuable to customers and strengthen the economics of delivering it. The right application will depend on the business.
How we’re living it
Building applied AI with management
We work with management across talent, commercial development, technology, and M&A. Applied AI runs through those functions: the skillsets a team needs, the services it can offer, and how it delivers them. Teams need to understand where the technology can help, where judgment remains essential, and how their work will change.
To turn that thinking into practice, we wanted someone who would work shoulder-to-shoulder with our teams and the businesses we back. That’s why Manoj Kothiyal joined Astira as Partner, Head of AI. Under his leadership, we are building our applied AI capability across the firm and portfolio. Manoj evaluates opportunities and risks with investment and management teams and helps develop detailed, multi-year roadmaps. The people delivering the service help shape the priorities.
The work starts with how a service is delivered: the information it uses, the decisions people make, and the exceptions they need to resolve. We examine where that process should change and where applied AI can help.
Astira AI Cortex helps accelerate this work through shared infrastructure, applications and expertise. It supports the firm-level capability we are building under Manoj’s leadership, with applications adapted to each business. A tool for reading financial documents may be reusable. Applying it well requires understanding the decisions it must support and the level of review those decisions require.
Building value through ownership
A roadmap helps management and investors decide where to commit resources. Across our four portfolio companies, we have mapped workflows, agreed on AI roadmaps with management, and launched selected builds. The first 90 days focus that work on understanding the process, setting priorities, and starting implementation.
Making applied AI work means changing how the work gets done. Teams need to know how to use the output, when to apply their judgment and where to direct the capacity created. Time saved creates value when it helps the business serve customers better or take on more work.
Once an application goes live, we ask:
- Are people using it?
- Does it work reliably?
- Does the benefit justify the cost?
We assess both financial returns and improvements in the service. Financial returns may come from lower delivery costs, additional work or stronger earnings, after accounting for implementation and ongoing operating costs. Better quality, faster responses, and a more strategic customer relationship also matter, even when their financial effect takes longer to establish.
Productivity, service quality, growth and financial results each tell us something different. Understanding how they connect is what guides where we invest next.
What this looks like in practice
Altus and Numera show how this approach applies to different specialist services. In both businesses, we invested in people and technology to change how the work is delivered and to give specialists more capacity to apply their expertise.
Altus frees collectors to focus on recovery
Altus provides commercial accounts receivable management services. Manual intake, data entry, letters, and legal forwarding constrain collector capacity and don’t deliver any incremental value for clients. Working with management, we added technology leadership and advisors to map workflows and developed a two-year AI roadmap.
AI agents now support these repeatable, low value tasks, while recovery models predict payment amounts, timing and legal outcomes to inform how accounts are handled. Collectors have more capacity for work that requires their experience and that improves outcomes for their clients. Outcomes are captured to improve the models over time.
Supported by AI-enabled workflows, Altus reported 10% year-on-year placement growth in the first half of 2026 without increasing collector headcount.
Numera creates more capacity for expert work
Numera provides outsourced accounting, tax, and CFO advisory services, and the company is rethinking accounting from the transaction level up. We worked with the company to stand up an in-house AI and engineering team and a cross-functional leadership council to support the work. The work sits within a 12-month AI roadmap already in execution. It changes both how transactions are processed and where professionals apply their judgment.
One of the first initiatives covered bookkeeping automation. Its bookkeeping professionals spend meaningful time on transaction posting and repeatable close tasks. A multilayered model combines applied AI with human review and a continuous learning loop. Experts’ corrections guide subsequent work, helping reduce the number of postings that require human intervention as the system improves.
The bookkeeping automation tool now consistently posts transactions accurately, with human override needed only 16% of the time. For clients, accuracy is central to the service. Numera’s multi-layer review process combines AI with human judgment to improve precision. As the tools improve, professionals can focus more of their judgment on resolving exceptions, reviewing outcomes and providing clients with better insights and advice.
The questions shaping how we invest and how we build
Our work at Altus and Numera brings wider investment and operating questions into focus:
- As applied AI expands what specialists can do, how should their roles evolve to deliver more value to customers?
- When applied AI raises productivity, what determines how the gains are shared between customers and the business?
- As applied AI takes over routine work, where will junior professionals learn the judgment customers rely on?
- How will the market value the work we do to modernize these businesses and prepare them for the future?
If you’re tackling some of these same questions in your business, we would love to hear from you.