Live lab instrument integration.
Connect scientific instruments directly to Draftbox. Watch measurements stream live in your browser. Annotate anomalies as they happen. Ask Aura AI about the data while it's still coming in. Your whole lab sees the same stream — simultaneously. This is not export-and-upload. This is real-time. Available exclusively on Scientist Lab.
Real-time streaming
Instrument measurements stream live. Multiple researchers can watch the same data simultaneously — no waiting for exports, no manual uploads.
AI analysis of live data
Ask Aura AI about live instrument data and get grounded answers with provenance back to specific data points. Not summaries. Analysis.
Concurrent annotation
Read and annotate instrument streams while they're live. Multiple researchers can mark anomalies, add notes, and flag data points — all while the instrument continues streaming.
Scientist Lab — what's included
| Instruments included | 1 (add more at €1,999/instrument/month) |
| Measurements per instrument | 10 million / month |
| Concurrent live streams | 10 per instrument |
| Seats | Unlimited — whole lab shares |
| AI analysis | Full reasoning verification on live data |
| Collaboration | Live concurrent annotations + real-time updates |
| SLA | 99.5% uptime guarantee |
What it looks like in practice.
Not features. Not specs. Here is what happens when your lab runs on Draftbox.
HPLC run — 3 researchers, 1 stream
Your HPLC starts a 40-minute run. The chromatogram streams live into a Draftbox document. Three researchers open the same document — one in Athens, one in Munich, one in Cambridge. They all see peaks appear in real-time. One marks an anomaly at minute 12. Another adds a note: "check retention time against standard B." The third asks Aura: "Is this peak consistent with last week's run?" — and gets a grounded answer with provenance to both runs, before the instrument finishes.
Mass spec anomaly — caught while you wait
A mass spectrometer streams spectra into Draftbox. Aura AI watches the incoming data and flags an unexpected ion peak that does not match any reference in your corpus. The verification model confirms the anomaly is real — not noise. You get an alert in the document before the run completes. You annotate the peak, link it to a hypothesis, and share the annotated spectrum with your PI — all in the same document, all before the instrument cools down.
Multi-instrument experiment — one canvas
You run a kinetics experiment across three instruments: a spectrophotometer, a pH meter, and a temperature logger. All three stream into the same Draftbox document. KaTeX equations define the rate law. Mermaid diagrams map the reaction pathway. Live data blocks show absorbance, pH, and temperature side by side — updating every second. Aura answers "is the rate consistent with Arrhenius at this temperature range?" by reading all three streams simultaneously. Your lab sees the answer. The instrument keeps running.
The gap between your data and your conclusions.
Research generates mountains of data — papers, protocols, measurements, instrument outputs. The tools you have either store it or summarize it. Neither is enough. You need something that reads it, understands it, verifies it, and connects it — with provenance, in real-time. Not after the experiment. During it.
Finding what matters
10,000 pages of notes, 500 papers, 50 instrument logs. Which paragraph answers your question? Draftbox finds meaning, not just keywords — and every answer traces back to the exact source.
Trusting the answer
An AI that summarizes is a starting point. An AI that verifies its own reasoning — a second model checking the first — is a collaborator. Every complex answer is independently checked before you see it.
Connecting the lab
Your instruments produce data continuously. Your ELN should consume it continuously. Scientist Lab connects instruments via API — live streaming, concurrent annotations, AI analysis of measurements as they arrive.
Two models. One verified answer.
When Aura AI generates a complex research answer, a second independent model checks the reasoning for logical errors, hallucinations, and gaps. If the verification model disagrees, the answer is flagged before you see it. This is not a self-graded exam. It is peer review, automated, on every complex query — while you wait.
Deep reasoning
Multi-step analysis across your entire corpus — entire papers in a single reasoning chain
Independent verification
A separate model architecture checks each reasoning step for consistency and source alignment
Full provenance
Every claim traces back to the exact block in your data — click to see the source, in context
A canvas built for research.
Not a text editor with plugins. Draftbox is a block-level canvas with infinite pages — write equations in KaTeX, draw diagrams in Excalidraw, embed live data blocks that poll external sources, render Mermaid graphs, and organize everything in multi-column layouts. All in one document, all searchable by AI.
KaTeX equations
Full LaTeX math rendering inline and in blocks
Live data blocks
Embed HTML blocks that poll external data sources in real-time
Mermaid diagrams
Flowcharts, sequence diagrams, state machines — rendered inline
Excalidraw whiteboard
Sketch, annotate, and draw directly inside your documents
Infinite pages
Block-based architecture — documents scale without limits
29 integrations
Sync from Linear, Jira, GitHub, Notion, Google Drive, and 24 more
Fast enough for real research.
Scientist plans process documents in parallel, cache answers before you ask, and prioritize your queries over everyone else's. The result: research-scale throughput without research-scale waiting.
~30s
To ingest 1,000 pages
64
Parallel workers on Scientist Lab
~50ms
Response time on cached queries
14 days
Cache retention on Scientist Lab
Scientist vs. standard plans — the difference
| Capability | Business (€79) | Scientist (€249) | Scientist Lab (€4,999) |
|---|---|---|---|
| AI Model (deep) | Claude Sonnet | Claude Sonnet | Claude Opus 4 |
| AI Model (balanced) | Gemini Pro | Gemini Pro | Gemini Pro |
| Reasoning steps | 8 | 14 | 14 |
| Parallel workers | 8 | 32 | 64 |
| Lift Power | 5M | 50M | 100M |
| Cache retention | 24h | 7 days | 14 days |
| Verified answers | — | ||
| Priority dispatch | — | ||
| Cache warming | — | ||
| OCR + voice | — | ||
| Instrument API | — | — | |
| Storage | 20GB | 100GB | 500GB |
| SLA | — | — | 99.5% |
Built to the standards your lab answers to.
The infrastructure aligns with the frameworks that govern scientific research — from lab competence to data integrity, from AI governance to privacy.
Lab & Research
- ISO 17025 (Lab competence)
- ISO 15189 (Medical labs)
- GLP (OECD principles)
- FAIR Data Principles
AI Governance
- ISO 42001 (AI Mgmt)
- EU AI Act
- NIST AI RMF
- OWASP Top 10 (LLM)
Data Protection
- GDPR (EU)
- HIPAA / HITECH
- ISO 27799 (Health)
- ISO 27701 / 27018
Security
- ISO 27001
- SOC 2 Type II ready
- FIPS 140-2 Level 2
- NIST CSF 2.0
No servers. No IT team. No maintenance.
Draftbox runs entirely in the cloud. No on-premise servers to buy, no databases to maintain, no IT staff to hire. Your lab connects through a browser — instruments through an API key. That's it. What you pay for is not infrastructure. You pay for a platform that watches your instruments in real-time, verifies every complex answer with a second AI model, lets your whole lab annotate live data simultaneously, and encrypts everything to FIPS 140-2 standards. That is the value. The infrastructure is free.
Zero on-premise infrastructure
No servers, no databases, no VPNs. Draftbox is hosted in EU data centers (Frankfurt) with edge functions deployed globally for low-latency access. Open a browser, connect your instruments, and your lab is live. That's the setup.
Your data is encrypted — and never used for training
Every field is encrypted at rest with FIPS 140-2 Level 2 hardware keys (AWS KMS). Data is hosted in the EU under GDPR. Your research data is never used to train AI models — not ours, not anyone's. If someone breaches the server, they see ciphertext. Breach detection runs every 30 seconds with automatic lockdown — your research stays hidden even under attack.
What you don't need
What €60K/year buys. And what it replaces.
Scientist Lab is priced as a direct competitor to enterprise lab software — because it does more, not less. Here is what you would pay for separately, and what Draftbox replaces in one platform.
| Capability | Benchling | Dotmatics | Draftbox Scientist Lab |
|---|---|---|---|
| ELN + notes | Yes | Yes | |
| Instrument connect | Module ($$$) | Module ($$$) | Included (1 instrument) |
| AI analysis | No | Limited | Full Aura AI + Opus 4 for complex queries |
| Live streaming | No | No | Yes — real-time |
| Concurrent annotations | Limited | Limited | Unlimited |
| Equations & graphs | Basic | Basic | KaTeX + Mermaid + Excalidraw |
| SSO + SLA | Yes (enterprise) | Yes (enterprise) | Included |
| On-premise infrastructure | Optional | Optional | None — fully cloud-hosted |
| IT staff required | Yes | Yes | No |
| Data used for AI training | Unclear | No | Never — encrypted, isolated |
| EU data residency | Add-on | Add-on | Included (Frankfurt) |
| Annual cost (1 instrument) | ~$50K | ~$40K | €59,988 |
Draftbox Scientist Lab is not cheaper than the alternatives. It is more expensive and more capable. The value is not "the same thing for less" — it is things that are impossible with other tools. AI analysis of live instrument data. Verified answers on every complex query. A canvas that handles equations, graphs, and real-time data. That is the difference.
Three tiers. One trajectory.
From solo researcher to connected lab. Each tier is a superset of the one below — no feature is removed, only added.
€249 / month
For the solo researcher processing large corpora.
- Verified AI answers (Gemini Pro → Sonnet)
- Priority dispatch — your queries go first
- Pre-computed cache — first query is instant
- Full-precision search across your corpus
- OCR + voice transcription
- 100GB storage, 50 workspaces
- KaTeX, Mermaid, Excalidraw, live data blocks
€499 / month + €199/seat
For research teams that share knowledge.
- Everything in Scientist, plus:
- SSO + SLA 99.5%
- Team analytics dashboard
- Shared cache across all team members
- 50 workspaces, 500GB storage
- Starts from 5 seats
€4,999 / month
For labs with connected instruments. Unlimited seats.
- Everything in Scientist Team, plus:
- Claude Opus 4 for complex queries (Gemini Pro → Opus)
- 1 lab instrument included
- +€1,999/month per additional instrument
- Unlimited seats — whole lab shares
- 10M measurements / instrument / month
- Live streaming + concurrent annotations
- AI analysis of live instrument data
- SSO + SLA 99.5% + team analytics
If the server is breached, your research stays hidden.
Every piece of content — every note, every measurement, every draft — is encrypted at rest. The system detects breaches in 30 seconds and locks the vault automatically. Your research is protected to the standards your lab answers to.
Encryption at rest
AES-256 with field-level keys — each field encrypted individually
Breach detection + auto-lockdown
30-second detection, automatic vault seal — data preserved, not destroyed
Behavioral authentication
The system learns your patterns — and challenges anyone who doesn't match them
Your research deserves infrastructure, not a chatbot.
Verified answers. Live instrument integration. A canvas built for equations, graphs, and real-time data. This is not a feature list — it is a research platform.