Feb 18, 2025
6 Min
Aistra team
They say that in every adversity there is an opportunity. For future job seekers, Agentic AI disruption being engineered by Silicon Valley is one such adversity. Y Combinator’s latest cohort funding a new generation of AI startup has some even pitching the end of vast swathes of current jobs. It’s no surprise that headlines are declaring an era of mass job displacement. Goldman Sachs recently projected that AI automation could replace up to 300 million full-time jobs worldwide, while the World Economic Forum predicted that 85 million jobs may be displaced by 2025.
AI Agents are set to handle a lot of routine and entry-level tasks—whether it’s drafting legal contracts, writing code, crunching financial numbers, catering to customer requests in contact centres, or even performing sophisticated market research and analysis. Several top business schools have reported a dip in their three-month post-graduation placement rates, indicating that even highly-skilled professionals are feeling the ripple effects of AI as we speak.
However, it’s not all gloom and doom. Aistra’s engagements with B2B enterprises reveal that the need for entire new categories of employment are emerging—demanding skill sets that blend technical, analytical, and creative thinking. Unlike a single AI model churning away in isolation, organizations are increasingly deploying multi-agent systems, a “digital workforce” of specialized AI agents. But will they have the same access to documented and undocumented data that humans have, will they adhere to standard operating procedures, will they adapt to a situational change, will their errors be covered by insurance or excused by ‘bosses’, will their success be impeded by their human co-workers fearful of losing their jobs, these are questions that will only resolve in time.
Here’s the picture we see in terms of the new age a team of AI Integrators that complement the AI Tech teams to provide oversight and critical thinking to ensure everything runs smoothly and ethically.
Within these multi-agent setups, a handful of new roles are rising to prominence:
Agent Managers: Think of these professionals as AI product managers. They design and define how different AI agents work together, specifying each agent’s “job description” and ensuring the overall agentic system aligns with business objectives.
Orchs (Orchestrators): These are the process gurus. They master the playbook, SOPs, and workflows so that each AI agent—whether it’s analysing data or handling customer inquiries—integrates seamlessly into customer operations.
Evals (Evaluators): Obsessed with quality, these folks make sure every AI agent performs to standard. They constantly test for accuracy, fairness, and reliability, catching issues before they escalate into bigger problems.
Data Ninjas: Responsible for the underlying infrastructure and data pipelines. If data is the lifeblood of AI, Data Ninjas ensure it’s clean, secure, and readily available to all agents that need it.
Over the course of this six-part blog series, we’ll break down the rapidly changing AI landscape—starting with a closer look at multi-agent architecture, then diving deep into the distinct roles that power it and share our experiences building and deploying 30+ AI agents in enterprises. By the end, you’ll see that while AI may well replace some jobs, it’s also creating a wealth of new opportunities for anyone ready to adapt. Rather than framing it as humans versus machines, it’s more productive to envision a partnership—one where professionals who embrace these new roles can future-proof their careers in an AI-powered world.
