Builder of AI-native systems that orchestrate real‑world behavior at scale.
Co-founder & CPO of Aktana, acquired by PharmaForceIQ in 2025, where he is now Head of Product. Fifteen years building AI products that move enterprise behavior, and named to the PharmaVoice 100 Most Inspiring in Life Sciences.
As AI saturates every workflow, the durable advantage is not the model itself but the structure, continuity, and governance of the context the model reasons over.
Ask Derek
Grounded answers, cited from Derek's real work.
Try:“How did Derek pioneer Next-Best-Action in life sciences?”
Systems shipped, behavior moved.
Pioneering Next-Best-Action
How AI replaced static segmentation in life sciences
Tactic Genie
Strategy planning as a generative search problem
Share of Answer
Category creationThe shift PFIQ is solving, and the operating model pharma hasn't built yet
Optipresent
From data → understanding → decision → impact
AI4Kidz
The same flywheel (engagement, data, insight, decision), kid-shaped
AI-Q / Hone
New ventureThe judgment layer: every AI eval scores the model, AI-Q scores the human
Three eras, one thesis.
Building next-best-action systems
Derek founded Aktana and designed the first NBA platform purpose-built for life sciences engagement.
Scaling AI adoption globally
Aktana reached more than 50% of the top 20 pharma companies and drove $1B+ in script lift, and Derek was named to the PharmaVoice 100.
From context to action
Now Head of Product at PharmaForceIQ, following Aktana's acquisition, and working one chain: a governed source of truth, the answer it earns, the judgment that answer has to survive, and the action that follows. Share of Answer and AI-Q sit at either end of it.
Notes from the field.
The Share of Answer: A Graphic Novel for the Age of AI Answers
A ten-chapter pharma-noir telling of the Share of Answer thesis. For decades the rules were reach, impressions, and share of voice, until one question changed the battlefield: "What treatment should I consider next?" The novel walks through the Answer Machine's hidden supply chain, the operating model shift from campaigns to questions, the invisible metric, the Context War, and the three cities of influence: Earned (trust), Paid (amplification), and Owned (context). It closes with the Share of Answer operating system. Read it on-site or download the PDF.
What's Your AI-Q?
Part 2 of the Judgment Gap series: a 0–100 measure of your judgment over AI-assisted work, on the premise that you can't improve what you can't measure. Derek makes the reviewer's implicit process explicit as six dimensions: Understanding, Verification, Assumption Awareness, Risk Recognition, Confidence Calibration, and Accountability. The structure is consistent enough to hold across coding reviews, strategy reviews, clinical reviews, financial reviews, and agent loops.
Share of Answer Isn't Just a Marketing Problem. Here's What Each Function Has to Build.
The operational follow-up to the Share of Answer operating model: what brand and commercial build differently, what's new for medical affairs and MSL teams, what MLR and regulatory have to monitor that they weren't monitoring before, and how the HCP journey changes when AI engines become the first touchpoint in a decision the brand never sees. It is framed by the Pew finding that only ~1% of users click a source link inside an AI summary, which makes the reply itself the impression.
The AI Judgment Gap
Why the next decade won't be won by the people with the best AI, but by the people who can still tell when it's wrong. The essay opens with the room that went quiet, a polished recommendation whose presenter couldn't answer "why this approach and not the other one?", and it pairs that with Derek's own story of confidently citing a statistic he later couldn't find. The growing distance between what we ship with AI and what we actually understand is the Judgment Gap.
From Share of Voice to Share of Answer: The Operating Model Pharma Hasn't Built Yet
Derek's flagship Share of Answer thesis, worked out for pharma commercial. He marshals 2026 adoption data (81% of US physicians now using AI in clinical practice, and generative AI overtaking sales reps as a clinical information source) to argue that the contest has shifted from share of voice and share of mind to Share of Answer: being the structured, trusted source cited inside the one synthesized answer an HCP receives. He then names the three structural shifts to the commercial operating model that most pharma teams haven't started.
Paid Media on AI Answer Engines: Unavoidable, Expensive, and Easy to Waste
The paid-media companion to the Share of Answer thesis. Anchored on a June 2026 JAMA Viewpoint warning that sponsored summaries inside AI clinical-decision tools like OpenEvidence read to physicians as evidence rather than advertising, Derek argues that paid placement in answer engines is now unavoidable and under regulatory scrutiny, and he then lays out how to spend smart instead of the way most pharma teams are about to waste budget.
Harnessing AI to Transform End-to-End Customer Engagement for Pharma and Biotech
Derek's case for moving past point solutions to a single end-to-end AI system for biopharma commercial. Nearly 6 in 10 pharma leaders report 2x ROI from AI within a year, yet under 5% consider their organization mature, and the gap is scaling across decentralized teams. He argues that a unified end-to-end system can compress commercial planning and execution from roughly 18 months to as little as six, unify brand and field, refine strategy and tactics in real time, and measure impact on business outcomes like script lift as it happens, so that hyper-personalized, adaptive HCP journeys become achievable at scale.
Why Context-Aware, Learning AI is the Future of Life Sciences
Derek's case for moving life sciences AI from static models to real-time learning agents. He argues that the unit of value is a centralized, dynamic knowledge base that every agent reads from and writes back to, and that single-agent systems will give way to coordinated multi-agent systems across HCP engagement, content, and resource allocation.
Transparency Is Non-Negotiable in AI
An argument that traceability is a product requirement, not an ethics talking point. Derek anchors it in an early-deployment story where field reps rejected unexplained recommendations until the system surfaced its reasoning, and that lesson became foundational to how Aktana's NBA was designed.
Future-Proofing CRM Investments in Life Sciences
Three principles for life sciences teams working through the Salesforce Life Sciences Cloud and Veeva Vault CRM transition: lead with a future-focused vision, abstract above any single CRM platform, and tie every migration decision to a measurable business outcome.
Harnessing AI for Biopharma Customer Engagement
A joint Q&A with PFIQ CEO Hemal Somaiya covering the PharmaForceIQ-Aktana combination, the move from omnichannel to optichannel, and how to balance AI-driven personalization with the human relationships that still close pharma deals.
Fine-Tuning, Prompt Engineering: the Keys to Real GenAI in Pharma
How fine-tuning and prompt engineering let pharma teams adapt general-purpose LLMs into trustworthy commercial tools for personalized content generation, automated tagging, and extraction of structured insight from unstructured HCP interactions.
AI 2.0: The Humanizing of Machine Learning Technology
Derek's framing of "AI 2.0," which is machine learning combined with business logic and human insight to produce contextual recommendations a rep or MSL will actually act on. He wrote it as commercial pharma absorbed the post-COVID shift to digital-led HCP engagement.
Press, quotes, and recognition.
Trajectory in one page.
Two decades across consulting, strategy, and AI-native product. The throughline is the same problem showing up in different surfaces: how to orchestrate real-world behavior over a shared, contextual substrate.
Derek Choy
AI-Native Product Executive · San Francisco Bay Area
University of Melbourne · B.Sc. Computer Science · LL.B (Honors) Law
PharmaVoice 100 Most Inspiring People in Life Sciences (2019) · Frequent contributor in industry trade press
Let's talk.
This is less an inbox than a doorway, so here is what's open and the cleanest way in.
Brainstorm & build
I like working with other product builders, AI engineers, and founders who are thinking through agentic systems, the AI context layer, and category creation. Trades and provocations are welcome.
Advise & be advised
Mentoring goes both ways. I am especially open to early-stage AI-native teams in life sciences and family software, and to learning from people building in adjacent domains.
Speak & on podcasts
I speak on agentic AI, NBA → NBE, attention engineering, AI in regulated industries, and the AI4Kidz thesis, and I am happy to keynote, sit on a panel, or come on as a guest.
PharmaForceIQ inquiries
If your team wants to see the Share of Answer offering, which is a governed source-of-truth context layer served to AI answer engines, or optichannel orchestration in action, I'll connect you with the right people at PFIQ.