Request a Consult

Request a Consultation

Your consultation request has been received
Oops! Something went wrong while submitting the form.
By hitting submit below, you consent to allow Fractional AI to process and store the data above. Please view our Privacy Policy for more detail.

Why Fractional AI? Answering the question I ask the most. 

November 12, 2025

In all 67 interviews I did last month, I asked some version of: “...so why are you taking the call?” or “what interests you about Fractional AI?”

Everyone had their own motivators — mission, people, learning, stability, perks.

Not every company is a fit for every person. The best ones are clear about what they’re trying to be the best in the world at, and equally clear about what they’re not. They double down on what makes them distinctive rather than trying to please everyone.

There are a lot of reasons someone might be excited about Fractional AI:

  • Incredible team of 35 (and growing). 60% of our engineering team are former founders, 100% have 10+ years of experience, 60% have PhDs or Masters. This means exceptional colleagues (and great chess tournaments).

  • Alignment with our thesis. We believe most of the value from gen AI will go to legacy companies sitting on previously unsolvable automation challenges, and that solving these problems requires custom solutions, not a one-size-fits-all SaaS product.

  • Veteran founders. We have repeat founders who’ve been building together for over a decade (with a few exits along the way - see Datavant, LiveRamp (NYSE: RAMP)).

  • Competitive compensation. We pay competitively, make everyone an owner (with a 10-year PTE window — no golden handcuffs), cover 99% of premiums, and offer a 401k match.

All valid reasons. But not the singular reason.

Where we truly win: Fractional AI is the best place in the world to learn applied AI.

Optimizing for Learning Applied AI

For candidates hungry to learn applied AI, I believe we’re the best company in the world. That comes down to three things:

  1. Our business model
  2. Our rituals
  3. Our pace

1. Our Business Model → The Most AI Projects per Capita

Our business model forces you to stay on the cutting edge of applied AI. You can’t help it.

Top companies come to us with problems that were previously impossible to solve without gen AI, whether that means building a voice agent to automate research interviews, a safety agent for a rail logistics company, or an AI API integration builder for SaaS business.

While most PMs or engineers might build one AI product a year (if that), at Fractional AI you build one per quarter, tackling the toughest, most important AI challenges our customers face.

Take one engineer’s first 16 months:

  • Built an AI content moderation agent for Change.org
  • Built an AI-powered integrations builder for Zapier
  • Built an AI system to automate BPO processes for an e-commerce company
  • Built an AI customer onboarder for a B2B SaaS company in the GRC space
  • Built an AI product for ADP

That volume of real-world reps leads to unmatched pattern recognition, intuition, and a growing library of applied-AI techniques which are key to building systems that really work.

Our engineers defeated bias in AI research by developing a "debating agents" framework, improved a voice agent's instruction following with a novel adaptation of chain-of-thought, and so much more.

You develop the kind of instinct that only comes from doing, not reading or theorizing. You can watch people debate RAG vs finetuning on Twitter as much as you want but building actual products requires novel insights from many projects.

2. Our Rituals → Collective Expertise Becomes Individual Expertise

The real multiplier comes from taking the ceiling off of individual exposure and empowering everyone with the collective intelligence of everything Fractional AI has built and learned.

At most places, your AI expertise = the sum of your personal projects.
At Fractional AI, your AI expertise = the sum of all Fractional AI projects.

How we make that real:

  • Apprenticeships. Every new engineer starts apprenticing on a real client project. They own real features (not toy projects) while learning the tools and frameworks of applied AI.

  • Tuesday Toolsday. Every Tuesday at lunch, the team shares new AI tools or techniques. Recent sessions have covered prompt autotuning and guest speakers from companies like Arcade AI.

  • Thursday Project Shareout. Every Thursday, one team presents a current client project architecture and demo (anonymized as needed), and the whole org shares feedback and ideas.

  • Bench time. Between client builds, engineers have the autonomy to take on projects that contribute to Fractional (think Google 20% time + AI). Past projects include a voice agent that automates post-project documentation and experiments investigating data formats and LLM processing.

  • Collaborations with OpenAI. We regularly collaborate with foundational model teams, provide feedback, and co-develop playbooks (see our Cookbook with OpenAI).

  • AI paper series. Every month we read and meet to discuss recent journal publications. Last month we read this one (“Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models”).

  • AI Coffeehouse. Quarterly, we host a practitioner-only meetup in our SF office — a space for demos, discussion, and cross-pollination.

Most of these rituals were conceived of by Fractional engineers, with edits and additions happening in real time.

3. Our Pace → Growth Compounds

In 18 months, we’ve gone from a 6-person WeWork closet (...I mean office) to:

  • A team of 35+ (and growing)
  • A $12M+ annual revenue run rate
  • A waitlist of customers
  • 30+ clients served
  • 100% of business coming from word of mouth and referrals

When a business grows revenue 6x year-over-year, the company is different every quarter, if not every month. New functions, new products, new geographies – so much greenspace to build.

Knowing What We Are (and What We’re Not)

Just as important as what we are: what we’re not

We’re not a chill job — we have high standards and strict client deadlines. We’re not a perks-first culture — there’s no onsite barista (though there’s a live contest for an espresso machine).

These are reasonable things for people to care about; they just aren’t what we care about.

Part of asking “Why Fractional AI?” is to better understand what candidates are looking for too, so we can help them opt in or opt out. 

But if you want to learn applied AI — really learn it, by building, shipping, and growing alongside elite peers — I can think of no better place.

Special Thanks to Holly May - a mentor, leader, and friend who’s “Why Datavant” post inspired this rendition, and also convinced me to sign up for my first startup journey.

Annie Powers is the VP of People and Operations at Fractional AI. Before Fractional, she led the People Team at Datavant and was an Economic Consultant at Analysis Group.

Explore other blog posts

see all