Strategic Intelligence Briefing  ·  Prepared for the Century Health leadership teamIndependent research by BioCreative Strategies
BioCreative Strategies
For Vish Srivastava and Lisa Johnstone & the Century Health leadership team

From 97% CHARM accuracy to a repeatable pharma sales engine — built in three months.

A custom React app, your own SQL database, a knowledge graph of every pharma medical affairs lead, HEOR buyer, and RWD decision-maker across top-20 and mid-size pharma — multi-channel outreach across LinkedIn and email — and the infrastructure that turns Century's validated data platform into a full commercial pipeline.

Everything on this page is built specifically for Century Health. The intelligence is live. The infrastructure is ready to deploy.

The full BioCreative engagement

Three waves. One commercial engine. Yours at the end.

The BioCreative Launch program builds Century's complete commercial system in three overlapping waves — pharma buyer intelligence first, then the database + UX + sending infrastructure, then live activation across LinkedIn and email. The output isn't a lead list — it's the operating system your 2-person commercial team scales with.

Wave A — Month 1

Intelligence + pharma buyer mapping

Deep market research, buyer universe mapping across top-20 pharma (Medical Affairs, HEOR, RWD/RWE, Commercial Analytics buyers), mid-size specialty pharma (faster decision cycles), and academic medical centers with pharma partnerships. Competitive positioning vs. Flatiron, Truveta, Aetion, OM1, Komodo, IQVIA. The knowledge graph populated with real signals — RWE publication activity, conference presentations, clinical trial data needs — from day one.

Wave B — Month 1–2

Database, UX, sending infrastructure

Postgres database seeded with the full buyer universe. Custom React dashboard deployed, sending domains warmed, HeyReach + EmailBison configured on dedicated IP. Built in Century's brand, behind your team's login. The commercial backend that walks Vish and Lisa into a live pipeline rather than a blank CRM.

Wave C — Month 2–3

Activation across LinkedIn + Email

Multi-channel outreach live targeting HEOR directors, VP Medical Affairs, RWD decision-makers, and commercial analytics leads at pharma companies actively investing in real-world evidence. Custom messages positioned around CHARM's 97% accuracy, community-practice data network, and disease-specific registry depth. Reply classification active. Pipeline reporting live. Meetings booked.

Every section below dives into a piece of this. What we build, how it works, and proof it's already running.

Your app build — what you own

The core stack we build, harden, and hand back.

A complete commercial system — custom React UX, Postgres database, backend automation, and hardened sending infrastructure — built in Century's brand. Cancel anytime and take everything with you: the app, the database, the server, the code. No lock-in, no black box.

Custom UX

Your React App

Accounts, campaigns, news, outreach — one operator UX in your brand, behind secure login. Hosted on your virtual server. Transferable to any host.

Database

Postgres / Supabase

SQL database. Real-time, exportable, API-accessible. Every table, every row — yours.

Hosting

Virtual Server

The backend hub for everything we build — APIs, automations, news ingestion, channel connections all run here. Available for migration if needed.

Code & IP

GitHub

Every line of code — front-end, back-end, schema, knowledge graphs, data pipelines, agent harnesses, prompts. All in your repo, all yours to transfer.

Backend

n8n + Custom Harnesses

Edge functions, coordinated n8n flows, custom Python and TypeScript harnesses — all executing on your server. The automation layer that keeps everything running.

LinkedIn

HeyReach

Multi-account LinkedIn automation. Your workspace, your rep accounts, your campaign data.

Email

EmailBison

Warmed sending infrastructure on a dedicated BioCreative IP. Your domains, your mailboxes.

Every piece above becomes Century's property at handoff. The UX, database, server, and GitHub repo form a transferable unit — move it anywhere, run it independently.

Our backend research engine — what powers your database

How we fill your knowledge graph.

These are the tools and feeds we operate continuously to populate Century's database with classified accounts, enriched contacts, scored news, and live intelligence. You don't license these individually — we run them and the structured output lands in your database, ready to power messages, the live newsfeed, and your team's daily UX. We've already used a combination of these tools to begin building this page and populating yours.

Clinical

ClinicalTrials.gov

Sponsor, site, PI, phase, indication — live trial graph.

Literature

PubMed · bioRxiv

Publication record, co-author graph, preprint signal.

Funding

NIH RePORTER

Active and historical grants, awards by lab and PI.

Regulatory

FDA + SEC + USPTO

Regulatory, capital markets, and IP signals.

Enrichment

Clay

Waterfall enrichment — emails, firmographics, technographics.

Enrichment

FullEnrich

Multi-source contact verification and enrichment cascade.

People

Sales Navigator

Title, tenure, company moves, intent signal.

Web intelligence

Firecrawl + Brave Search

Full-page scrapes of competitor sites, prospect pages, branding, and layout patterns — plus real-time web search for news, market signals, and public intelligence. Feeds knowledge graph continuously.

News feeds

RSS + News APIs

Self-contained pipeline: RSS feed scrapes + news API calls → keyword alignment against your watchlist → AI classification and per-client summaries. Runs 2×/day. Powers the live newsfeed in your UX.

Static LLMs we orchestrate

Claude · GPT · Gemini · Perplexity — these LLMs work as static processors. We engineer the context around them — what goes in determines what comes out.

Custom news intelligence — a self-contained pipeline feeding your knowledge graph

Scored news, in your UX, updated twice daily.

We build a custom news pipeline for each client: RSS feed scrapes + news API calls pull raw articles from dozens of sources, then custom code runs keyword alignment against Century's watchlist — your competitors, target accounts, therapeutic areas, regulatory bodies. Brave Search adds web-level enrichment where RSS gaps exist. An AI classifier scores each article for relevance and writes a per-client summary of why it matters to your business.

PRNewswire · May 2026

Century Health Closes Oversubscribed $5M Seed Round Led by Origin Ventures

Funding signal — outreach to pharma buyers can reference Century's fresh validation and growth trajectory. Board composition (Origin Ventures + Zorba Lieberman, Citeline founder) signals serious RWE market credibility. Use in outbound to establish institutional backing.

Nature Medicine · May 2026

FDA Signals Increased Acceptance of AI-Curated Real-World Data for Regulatory Submissions

Regulatory tailwind — pharma HEOR teams are gaining mandate to source AI-validated RWD for regulatory filings. Century's 97% CHARM accuracy vs. expert abstraction positions them squarely in this trend. Pre-outreach messaging can cite this shift to create urgency.

Two real items pulled and scored this week. The system runs continuously — you see every signal that matters, nothing that doesn't.

Proof we've already started populating your knowledge graph

We've already mapped Century's buyer landscape.

Research covering Century's competitive positioning, pharma buyer universe, and the gap between platform validation and repeatable commercial motion. Three highlights below.

Buyer access

14,800+ pharma contacts in our database — named RWE, HEOR, and Medical Affairs leaders at every top-20 pharma.

AstraZeneca alone: 14 RWE + 9 HEOR + 63 Medical Affairs titles in our active database. Amgen: 3 RWE + 9 HEOR + 35 MA titles. Across all top-20 pharma, we've mapped the specific buying roles that evaluate platforms like Century — VP HEOR, Director RWE Operations, Head of Medical Affairs Analytics. These aren't scraped lists — they're enriched, persona-classified, campaign-ready contact records.

Competitive positioning

Century differentiates on community-practice sourcing and LLM abstraction — but the market doesn't know yet.

Flatiron is oncology + academic. Truveta is broad-horizontal health systems. Komodo is claims-dominant. Century's multi-therapeutic disease registries from community practices fill a gap none of the incumbents cover — but with a 2-person commercial team, awareness hasn't scaled to match the platform's capabilities. The outbound system we build closes that awareness gap at scale.

Market timing

RWE budgets growing 14.8% CAGR — pharma is buying more real-world data, not less.

Even in a tight R&D environment, pharma real-world evidence spending is accelerating ($5.4B in 2025, projected $10.8B by 2030). Medical Affairs teams are gaining budget authority. HEOR is required for market-access submissions in every major geography. Century's timing is right — the question is whether the commercial infrastructure can match the market demand before competitors fill the gap.

Same knowledge graph, same data layer, same UX you'll log into. We've started; activation is the next step.

From raw data to outreach-ready contacts

Multi-level enrichment. Multi-source find-people. Classified for fit.

Every account and contact in Century's database runs through a layered enrichment pipeline before a single message gets generated. Each level is a deterministic gate — cost-controlled, source-traced, validated. Junk goes out; signal stays in.

Account track

Multi-level enrichment + ICP scoring

Accounts run through progressive enrichment levels — light enrichment fills missing data via search + LLM, deep enrichment runs category-aware research into structured intelligence fields. ICP scoring is the final gate: tier and fit score per account, so outreach only targets companies worth reaching.

Contact track

Find-people + enrichment cascade

Sales Navigator + PhantomBuster + Clay's multi-provider find-people waterfall + custom scrapes locate the right contacts at every targeted account. Then full-profile enrichment runs: headline, summary, publications, work history, verified email waterfall — everything needed to write a message that reads like a human did the research.

Classification

Marketing persona mapping

Every eligible contact gets a Claude-graded marketing persona classification for the SSO matrix. Persona tags drive message differentiation, channel prioritization, and sequence selection. Trigger detection — job changes, publications, funding events — populated continuously.

By the time the messaging engine runs, every contact already carries enrichment, fit score, and persona classification. The same tags power the SSO matrix and every channel downstream.

How every message gets written

SSO Matrix — a context engineering agent, not a prompt.

When it's time to write an email or LinkedIn message, we don't ask an LLM to “be creative.” We run the SSO Matrix — a context engineering agent that pulls everything relevant about this person, this account, this moment — and assembles a precise context window before the LLM ever sees the task.

Pain points and value props aligned to the ICP and buying persona · enrichment record (trials, publications, funding, news, role context) · recent triggers · message-thread history · brand voice tokens. All assembled, then handed to a static LLM for generation.

1
Pull
Enrichment record + ICP classification + persona + triggers + brand voice tokens
2
Assemble
Context engineered into a structured matrix — the SSO payload
3
Generate
Static LLM writes the best possible message for THIS person at THIS moment
4
Score + ship
Brand Voice Guardian + deliverability checks → send via HeyReach or EmailBison

Following Karpathy's method

The LLM is a static processor. The intelligence comes from the context engineering — what we pull, what we filter, what we hand it. All the work happens before the model sees the task. That's why every message reads like it was written by someone who knows the recipient's business.

Channel 1 — LinkedIn via HeyReach

Blank connection → accept → follow-up DM → reply handoff.

Multi-account LinkedIn automation running across multiple rep profiles. Connection requests, DM sequencing, acceptance tracking, reply classification — all orchestrated through HeyReach. Each rep account warms independently so no single profile gets throttled.

1
Blank connection
Clean request sent from rep's account. No note — higher accept rate.
2
Accept tracked
Webhook fires the moment they accept. Database updated in real-time.
3
Follow-up DM
SSO Matrix generates a message personalized to role, company, and enrichment data.
4
Reply → rep
Reply classified and handed to rep for human close. No AI replies without approval.

Generated examples — different persona, different message

VP HEOR
Director Medical Affairs

Dr. Sarah Mitchell, VP Health Economics & Outcomes Research at AstraZeneca

LinkedIn DM (after connection accepted)
Sarah — noticed AstraZeneca's expanded RWE investments in respiratory and immunology. We built an AI platform (CHARM) that delivers disease-specific RWD from community-practice EHRs — validated at 97% vs. expert chart abstraction. Curious if you're evaluating new data partners for the specialty registries your team is building.

Dr. James Okafor, Director Medical Affairs Analytics at Amgen

LinkedIn DM (after connection accepted)
James — saw Amgen's continued investment in real-world evidence for the nephrology and inflammation portfolio. We've built an AI platform that delivers disease-specific patient registries from community-practice EHRs — 97% accuracy on unstructured clinical notes. Could be relevant for the outcomes data your Medical Affairs team needs. Worth a quick conversation?

Why blank connections work

Connection requests without notes consistently outperform noted ones in accept rate. The follow-up DM — sent only after they accept — is where the personalization lives. By then they've already signaled intent.

What you own

HeyReach workspace, rep account access, campaign data, all sequence logic. Sales Navigator identifies the right contacts. BioCreative's classification agents ensure each connection hits the right persona at the right tier.

Channel 2 — Email via EmailBison

AI-generated sequences. Warmed inbox. Per-recipient personalization.

Purpose-built email sending infrastructure with dedicated warmed domains. Multi-step sequences, real-time tracking, reply webhooks, bounce handling. Every email lands in inbox because the infrastructure is engineered for it — not bolted on.

1
AI drafts sequence
SSO Matrix generates multi-step emails from knowledge graph + enrichment record. Per-persona, per-account.
2
Warmed inbox send
Delivered via verified sending infrastructure. DNS, DKIM, SPF, warm-up all handled.
3
Tracking
Opens, clicks, replies tracked in real-time. Engagement scoring updates the contact record.
4
Reply → rep
Reply webhook fires. Classified (interested / objection / referral / meeting). Routed to the right rep.

Generated examples — different persona, different message

VP HEOR
Director Medical Affairs

Dr. Sarah Mitchell, VP Health Economics & Outcomes Research at AstraZeneca

Email — Step 1
Subject: Disease-specific RWD from community practices — Century Health Dr. Mitchell, With AstraZeneca expanding its RWE capabilities in respiratory, nephrology, and immunology, I wanted to introduce Century Health. We've built an AI-powered platform (CHARM) that extracts clinical variables from unstructured EHR data — physician notes, radiology reports, pathology — at 97% accuracy vs. clinical expert abstraction. Unlike claims-based approaches, we deliver disease-specific registries with the clinical depth HEOR analyses actually require. Our data comes from community and specialty practices (not academic centers alone), giving you access to diverse patient populations that reflect real-world treatment patterns. We deliver in OMOP CDM format, in your preferred compute environment. Would a 15-minute conversation about how this maps to AstraZeneca's current RWE needs make sense? Best, [Rep name]
Email — Step 2
Subject: Re: Disease-specific RWD from community practices — Century Health Dr. Mitchell — brief follow-up. One specific data point: our data network grew 60x in the past year, and we're actively building registries in neurology, nephrology, respiratory, ophthalmology, and metabolic disease. For HEOR teams that need longitudinal real-world outcomes in specialty therapeutic areas — the clinical depth from unstructured notes is what makes the difference for regulatory-grade analyses. Happy to share a sample data dictionary or discuss a pilot scope — no commitment needed. Best, [Rep name]

Dr. James Okafor, Director Medical Affairs Analytics at Amgen

Email — Step 1
Subject: AI-powered RWE for Amgen's nephrology & inflammation portfolio Dr. Okafor, With Amgen's Medical Affairs team actively generating real-world evidence for the nephrology and inflammation portfolio, I wanted to introduce Century Health. Our platform (CHARM) uses proprietary AI/LLM technology to abstract clinical variables from unstructured physician notes — not just claims codes — delivering disease-specific registries with the depth Medical Affairs needs for publications, guideline support, and market access evidence. We specifically have nephrology as one of our anchor therapeutic areas, with established community-practice data sources that capture the specialty populations most relevant to Amgen's renal portfolio. Would it make sense to discuss how this aligns with your current evidence generation priorities? Best, [Rep name]
Email — Step 2
Subject: Re: AI-powered RWE for Amgen's nephrology & inflammation portfolio Dr. Okafor — brief follow-up. One angle that may resonate: community-practice data captures patients earlier in their disease journey and in more diverse care settings than academic-center-only approaches. For Medical Affairs publications and real-world evidence submissions, this broader population representation strengthens the external validity argument regulators and payers look for. Happy to walk through a sample nephrologyregistry or discuss pilot timelines if useful. Best, [Rep name]

Why we build separate sending domains

Protect your core domain

Century's primary domain reputation stays untouched. Outreach runs on dedicated domains we build, verify, and warm up over 2-4 weeks before a single cold email sends.

Full DNS + authentication

SPF, DKIM, DMARC, custom tracking domains — all configured from day one. Deliverability monitoring runs continuously. If a domain cools, we rotate.

Land in inbox, not spam

Gradual warm-up cadence, engagement-based send scheduling, bounce suppression, spam trap monitoring. Best practices baked into the infrastructure from day one.

Dedicated BioCreative sending IP

Why off-the-shelf senders can't match this.

Off-the-shelf senders (SmartLead, Lemlist, Instantly) route through shared IPs sending for hundreds of accounts. Microsoft and Google are increasingly throttling or blocking these IPs at the inbox boundary — placement collapses regardless of your domain reputation.

We send from a dedicated IP we own, used only for BioCreative clients, kept clean by tight quality control across every domain, mailbox, and message. Low-volume, warmed sends only. No links in cold messages. Every send vetted. That's why our deliverability holds where others' don't.

What you own: Sending domains, mailbox seats, all campaign data, sequence logic, reply history.

Two channels. Coordinated. One system.

LinkedIn + Email running off the same enrichment data, same classification, same knowledge graph. Multiple rep profiles per channel. One operator dashboard. Every touch coordinated so a prospect never gets conflicting signals — just consistent, intelligence-grounded outreach from Century's team.

How it works

Wave A → Wave B → Wave C. Three months to a running engine.

Each month maps to a wave. Intelligence first, then infrastructure, then activation. Every artifact yours to keep.

Month 1Wave A — Intelligence

Wave A — Intelligence + pharma buyer mapping

Deep research dossiers, buyer universe mapping across top-20 pharma, mid-size specialty pharma (the fast-close lane), and academic-pharma partnership networks. Competitive positioning vs. Flatiron/Truveta/Aetion/OM1/Komodo/IQVIA. Knowledge graph seeded with RWE publication activity, conference presentations, and pharma data procurement signals.

Delivered: Research dossier set, enriched account universe (14,800+ contacts), HEOR/MA/RWD buyer classification, ICP schema, knowledge graph live, pharma news scoring active.

You keep: All data, schema, query layer, refresh runbooks.

Month 2Wave B — Infrastructure

Wave B — Database + UX + sending infrastructure

Custom React app deployed — the commercial backend that walks Vish and Lisa into a live pipeline rather than a blank CRM. Account pages, campaign tooling, newsfeed, message drafting. Sending domains warmed; HeyReach + EmailBison configured.

Delivered: Custom React UX behind login, virtual server provisioned, GitHub repo populated, warmed sending domains, HeyReach workspace, EmailBison configured, dedicated IP active.

You keep: Hosting, code, sending domains, mailbox ownership, all workflow configurations.

Month 3Wave C — Activation

Wave C — Activation across LinkedIn + Email

Multi-channel outreach live targeting HEOR directors, VP Medical Affairs, RWD decision-makers, and commercial analytics leads. Custom messages per segment — top-20 pharma (long-game relationships), mid-size specialty pharma (fast-close opportunities), and academic-pharma partnership leads. Reply classification active. Pipeline reporting live. The 8–12 qualified discovery meetings in 90 days that actually get Century to the table.

Delivered: Live LinkedIn + email campaigns, AI-drafted sequences per persona, reply handoff workflow, pipeline dashboard updated, first meetings booked.

You keep: Campaign data, message history, agent prompts, sequence logic, reply records.

OngoingOperate

Optional: continuous tuning + conference activation + mid-size pharma fast-close lane

Quarterly tune-ups, pre-event intelligence sprints (DIA, ISPOR, AMCP), database refreshes, mid-size pharma POC acceleration, and the option to layer on provider-side parallel motion (expanding Century's data network through practitioner outreach).

Two support systems on top of your core launch

One going upstream into marketing. One going downstream into sales.

Both build on the same database, the same knowledge graph, the same brand system the core launch delivers. Both are scoped separately, activated when ready — not part of the core engagement, but built to plug straight into it.

Upstream — marketing channels that feed your funnel

Top-of-funnel amplification across every surface your buyers see.

Once the core engine is running, the same knowledge graph + brand system + database powers everything upstream — company social presence, executive thought leadership, search and AI-citation visibility, paid amplification, web optimization, and live calling. Marketing-owned, harder to track, but the connective tissue that makes outbound feel like a known voice.

We've already scraped Century's brand — colors, fonts, layout patterns — directly from century.health. This page is the proof. Every asset we produce downstream matches your visual identity without manual design work.

Company social presence

Content Command Center

LinkedIn company, X, Instagram, Threads, Bluesky, blog, newsletter. AI-generated social assets with image and video generation. News-driven concept generation matched back to Century's core thought principles. Brand Voice Guardian scores every output before it ships.

Tracked end-to-end: engagement on company posts feeds back into the knowledge graph as triggers and targeting signals.

Executive thought leadership

Scientific communications

Founder, scientific leader, and executive voice across LinkedIn personal, conferences, partnership outreach, investor updates, analyst notes. Same knowledge graph, same brand system — but a different voice and surface.

CCC connection: Engagement on personal posts is tracked at both the person level and the account level — engaged followers who match ICP become outreach targets automatically.

Search presence

SEO + GRO optimization

Google Analytics 4 + Search Console + AI citation tracking. Every asset keyworded and position-tracked. Century shows up in traditional search and in AI-generated answers (Generative Research Optimization).

CCC connection: blog posts, social assets, and backlinks from the Content Command Center compound into your search and AI citation presence.

Web presence

Website optimization

Using the same intelligence, brand system, and tools that built this page, we help Century create comprehensive, interactive, on-brand HTML assets — landing pages, product pages, microsites — informed by market data and kept up to date through our tooling.

Paid amplification

Paid channels

LinkedIn Ads, Google Ads, Meta Ads, programmatic display, retargeting — all powered by the same ICP classification and enrichment data. Target the exact accounts your outreach is hitting with awareness ads that reinforce the message across every surface.

Outbound calling

Cold calling — in partnership with Science2Sales

Dedicated calling team executing to high-value prospects identified by your enrichment pipeline. Direct conversations with people in real labs, real offices, real procurement teams. Market research, voice-of-customer intelligence, and live prospect feedback flow back into your knowledge graph alongside pipeline opportunities.

All built on the same knowledge graph, brand system, and database the core engagement delivers. Engaged after Wave C.

Downstream — strategic selling on the same back-end

Downstream support — strategic selling, powered by the same knowledge graph.

Once meetings start landing, the same back-end becomes the engine for managing pipeline, qualifying opportunities, and supporting your 2-person team in every pharma conversation. Built on Miller Heiman strategic selling — proven in complex enterprise health-data sales.

In core launch

Gold Sheets — account intelligence pages

The account-level page in your UX is the Gold Sheet: full enrichment, news, contacts (scored + persona-tagged), positioning notes, campaign history across LinkedIn and email — evergreen, updated as the world moves.

Downstream add-on

Pipeline tab — opportunities + multi-buyer mapping

When a Gold Sheet matures into a real opportunity (timeline, budget, qualification), it becomes a pipeline record. Map every buying-decision contact (HEOR budget owner / IT security reviewer / procurement / data governance), track interactions, and ingest meeting transcripts so the picture stays live.

Downstream add-on

Blue Sheets — meeting prep + in-meeting assets

Before every key pharma meeting, the system generates a Blue Sheet: who's in the room, their place in the buying decision, full prior-engagement history, mapped pain points, prepared objection responses, goals for the meeting, next-step recommendations.

Like the marketing channels above, these are downstream additions to the core launch — not separate builds. The infrastructure is already there; we activate it as your sales motion matures.

You own the system. Period.

Code, data, prompts, dashboards, infrastructure — all transferred to Century at handoff. We don't run an “AI black box” you keep paying us to operate. What we hand back is yours, the same way Century hands customers a real platform they own outcomes on.

Next step

Ready to see the full system.

Everything on this page is built specifically for Century. The intelligence is already underway. The infrastructure is ready to deploy. One conversation to scope it.

— Brian Elbert, BioCreative Strategies
brian@biocreativestrategies.com