Olto Discovery Platform

Olto Discovery is a research platform that combines AI-powered protocol generation with manual workflow design, simulation, and collaborative lab management. This guide covers everything you need to start designing research-grade experiments.

Account Setup

Create a free Explorer account at oltodiscovery.com. No credit card required. Confirm your email address to activate your account. You can add your institution, role, and research fields in Settings → Profile — this helps the AI assistant personalize its recommendations.

Your First Protocol

1. Navigate to New Protocol from the sidebar.
2. Enter your research goal in plain language — be as specific as possible. Include the target, model system, technique, and primary endpoint.
3. Select your scientific field and experiment type.
4. Check the equipment available in your lab.
5. Set your budget range and timeframe.
6. Click Generate Protocol. Olto will return a complete, research-grade experimental design in approximately 12 seconds.

Explorer Plan Limits

Free Explorer accounts can generate up to 5 protocols and 10 AI calls per month. Upgrade to Researcher ($79/month) for 500 protocols, 500 AI calls, the AI assistant, and literature intelligence.

The AI Protocol Generator converts a natural language research goal into a complete, structured experimental protocol. It is powered by Claude AI with field-specific scientific training.

Input Parameters

Protocol Sections

Every generated protocol includes ten structured sections:

1. Objective — A precise statement of what the experiment will determine or measure.
2. Hypothesis — A falsifiable prediction with direction and expected magnitude.
3. Experimental Design — Arms, n values, randomization, blinding, and primary endpoints.
4. Materials & Reagents — Specific reagents, cell lines, and consumables with catalogue numbers where applicable.
5. Procedure — Step-by-step protocol with concentrations, temperatures, and durations.
6. Variables — Independent, dependent, and controlled variables tabulated.
7. Control Groups — Positive controls, negative controls, housekeeping controls, and vehicle controls.
8. Expected Results — Predicted outcomes with statistical thresholds.
9. Quality & Reproducibility — QC metrics, acceptance criteria, and statistical analysis plan.
10. Safety & Risk Notes — Hazard classification, PPE requirements, waste disposal, and institutional approvals.

Quality Score

Every protocol receives four scores from 0–100:

• Feasibility — Likelihood of execution given budget, equipment, and timeframe constraints.
• Controls — Adequacy of experimental controls to support valid conclusions.
• Reproducibility — Specificity of procedural detail to enable replication.
• Clarity — Completeness and precision of instructions.

Protocol Refinements

Use the refinement buttons to iteratively improve any protocol without re-entering your parameters:

• Reduce cost — Substitutes expensive reagents and optimizes quantities.
• Compress timeline — Parallelizes steps and reduces incubation times where scientifically valid.
• Stronger controls — Adds negative, positive, isotype, and internal controls.
• Improve reproducibility — Adds QC checkpoints, tightens acceptance criteria, and specifies instrument settings.
• Add statistical plan — Includes power analysis, sample size calculations, and recommended statistical tests.
• Add troubleshooting — Documents 8–10 common failure modes with corrective actions.
• Simplify for uni lab — Adapts to equipment available in a standard academic setting.

The Manual Protocol Designer lets you build experimental workflows from scratch — step by step — with full control over every parameter. It complements AI generation for protocols that require proprietary methods, established SOPs, or highly customized workflows.

Accessing the Designer

Click Protocol Designer in the sidebar. You'll see three tabs at the top: Design, Test Run, and Simulation.

Protocol Metadata

Before adding steps, fill in the Protocol Overview at the top of the Design tab:

• Protocol title — Descriptive name for your protocol (editable in the toolbar).
• Scientific field — Select from the field dropdown in the toolbar.
• Objective — What does this protocol aim to demonstrate?
• Hypothesis — What do you predict will happen, and why?

Step Types

Each step has a type that defines its role in the workflow:

Step Fields

Click any step to expand its edit panel. Available fields per step:

• Title (required) — Brief, action-oriented name. e.g. "Wash cells with PBS × 3"
• Detailed Instructions — Full procedural text including volumes, concentrations, and technique notes.
• Duration — Numeric value with unit (seconds, minutes, hours, days). Used for timeline estimation and simulation.
• Temperature — In degrees Celsius. Leave blank if ambient.
• Expected Output — What should be visually or measurably observable after this step? e.g. "Cell pellet visible" or "OD600 = 0.6–0.8".
• Equipment — Select all instruments required for this step.
• Decision Branches (Decision Gate steps only) — Add labeled branches for each possible outcome and the corresponding next action.
• Step Notes — Tips, warnings, or context that do not belong in the procedure text.
• Critical Step — Toggle to flag steps where an error would invalidate the entire experiment. Critical steps are highlighted in red and emphasized during test runs.

Reordering Steps

Use the up/down arrow buttons on each step card to reorder. Steps are automatically re-indexed. You can also delete any step using the trash icon — you will not be prompted to confirm unless the step has significant detail filled in.

Saving to Library

Click Save to Library in the toolbar. Your protocol is saved with:

• All steps persisted to the database with full detail
• Auto-calculated quality scores based on step completeness, critical step count, and metadata richness
• Tagged as "manual-design" in your library for filtering

The simulation engine uses AI to analyze your protocol design before you run it in the lab. It identifies risks, estimates resource requirements, and provides optimization suggestions at the step level.

Running a Simulation

1. Open any protocol in the Manual Designer or click Simulate from the toolbar.
2. Click Run Simulation in the Simulation tab. Analysis typically takes 10–20 seconds.
3. Review the results across four output areas.

Simulation Outputs

Duration Estimate — Total protocol time calculated from step durations plus standard overhead (setup, cleanup, prep). Given in hours.

Cost Estimate — Rough reagent and consumable cost estimate based on step count, equipment, and field complexity.

Risk Flags — Specific issues identified at the protocol level, rated by severity:

• High — Issues likely to invalidate results or cause safety incidents. Must be resolved before running.
• Medium — Issues that may compromise data quality. Strongly recommended to address.
• Low — Minor gaps or opportunities for optimization.

Step-by-Step Analysis — For each step, the AI identifies specific issues (missing controls, temperature inconsistencies, ambiguous output criteria) and provides targeted improvement suggestions.

AI Recommendations — A 2–3 paragraph expert assessment of overall protocol feasibility, methodology rigor, and specific improvements to maximize success rate.

Simulation History

All simulation runs are saved and associated with the protocol. This allows you to track how protocol quality evolves across refinement iterations. Previous simulation results are accessible from the protocol detail page.

The Test Run tab provides a structured interface for walking through your protocol in real-time — either in the lab or as a dry-run review. Each step can be marked as Passed, Failed, or Skipped, with a timestamped run log generated automatically.

Running a Test

1. Ensure your protocol has at least one step in the Design tab.
2. Click the Test Run tab.
3. Work through each step in sequence. For each step:
   • Read the instructions and expected output.
   • Execute the step in the lab (or review it mentally).
   • Mark as Pass, Fail, or Skip.
4. Review the run log at the bottom of the page.

Step Status Indicators

• Pass — Step completed successfully with expected output observed. Marked in green.
• Fail — Step did not meet expected criteria. Record the specific deviation in your notebook.
• Skip — Step intentionally omitted for this run (e.g., not applicable to a particular experimental condition).

Critical Step Handling

Steps marked as Critical are displayed with a red left border and a warning indicator. A failure on a critical step typically means the experiment cannot be continued reliably and the entire run should be investigated before proceeding.

Run Log

The run log captures step name, outcome, and timestamp for every step actioned. This log can be copied and pasted into your Research Notebook as an experiment record. Future versions will automatically sync test run results to the notebook.

The AI Scientific Assistant is a persistent research intelligence that remembers your lab history, prior protocols, and preferences across every session. Available on Researcher and Lab plans.

What the Assistant Knows

The assistant has access to your full conversation history within each conversation thread. It is aware of your scientific field, role, and institution from your profile. In future versions, it will have access to your protocol library, uploaded papers, and notebook entries.

Capabilities

• Interpret experimental results and suggest next steps
• Recommend statistically appropriate analysis methods for your data
• Critique experimental design and identify confounds
• Explain scientific concepts at any depth level
• Summarize literature and identify research gaps
• Compare methodologies across your protocol history
• Generate draft materials and methods sections
• Troubleshoot failed experiments

Conversation Management

Each conversation is stored separately and accessible from the sidebar. Create a new conversation for each distinct research question or project. Keeping conversations focused improves the quality of responses. Use descriptive conversation titles for easy retrieval.

Rate Limits

Researcher plan: 100 messages per hour. Lab plan: 200 messages per hour. Enterprise: Configurable per organization.

The Research Notebook is a persistent, organized space for all lab observations, results, hypotheses, and protocol logs. All entries are timestamped, tagged, and associated with specific notebooks.

Notebooks

Create separate notebooks for each project, experiment series, or research area. Notebooks can be shared within your team (Lab plan) or kept private. Click New notebook in the sidebar of the Notebook page.

Entry Types

The Protocol Library stores all generated and manually designed protocols. Every protocol is versioned and preserved indefinitely.

Filtering & Search

Filter your library by scientific field using the tag pills at the top of the Library page. Filter to Starred protocols using the star filter. URL-based filtering means filters are shareable and bookmark-able.

Quality Scores in Library View

Each protocol card shows the overall quality score (average of four sub-scores) with a color-coded bar. Protocols scoring 85+ are shown in green, 70–84 in purple, 55–69 in amber, and below 55 in red.

Versioning

Every refinement creates a new version. The current version number is shown in the protocol detail view. Version history with before/after snapshots is recorded for all refinements.

Starring

Star important protocols for quick access. Starred protocols appear at the top of filtered views and in your dashboard. Click the star icon from any protocol card or detail view.

Team features are available on the Lab plan ($499/month) and Enterprise. Teams are organized into Organizations — each organization has its own member list, permissions, and shared workspace.

Creating an Organization

1. Navigate to Team in the sidebar.
2. Click Create team.
3. Enter a team name and optional description.
4. You are automatically added as the team Owner.

Member Roles

Inviting Members

Click Invite member on any team card and enter the email address. An invitation link is sent valid for 7 days. The invitee must have or create an Olto account to accept.

Protocol Approval Workflow

On Lab and Enterprise plans, protocols can be submitted for approval before they are marked as active. The approval status field tracks: Draft → Pending Review → Approved or Rejected. This supports institutional oversight requirements and good laboratory practice.

Upgrading

Go to Billing in the sidebar. Select your desired plan and click to proceed to the Stripe checkout. Changes take effect immediately upon payment confirmation.

Cancellation

Cancel anytime from Billing → Manage subscription. Your plan remains active until the end of the current billing period. No data is deleted upon cancellation — your library remains accessible on the Explorer free tier.

Academic Discounts

Verified academic institutions receive 40% off Researcher and Lab plans. Email academic@oltodiscovery.com from your institutional email address.

Olto Discovery is built with institutional security requirements as a baseline — not an afterthought. Research IP, protocol designs, and experimental data are protected by multiple independent security layers.

Encryption

• At rest — All data encrypted with AES-256 via Supabase's managed PostgreSQL infrastructure.
• In transit — All API traffic encrypted with TLS 1.3. HSTS enforced on all endpoints.

Data Isolation

Every database row is protected by PostgreSQL Row-Level Security (RLS) policies. Users can only read and write data they own. Organization members can access shared resources only within their organization scope. These policies are enforced at the database level — they cannot be bypassed through the application layer.

Audit Logging

Every significant action is logged to the audit_logs table with: user ID, action type, resource identifier, IP address, user agent, and timestamp. Logs are immutable (no update/delete policies). Lab and Enterprise plan administrators can view audit logs from the Team settings page.

Authentication

Passwords are hashed using bcrypt via Supabase Auth. Two-factor authentication is available and recommended for PI accounts. SSO/SAML integration is available on Enterprise plans for institutional identity providers.

SOC 2 & HIPAA

Olto's infrastructure is aligned with SOC 2 Type II controls. HIPAA-ready deployment options with a Business Associate Agreement (BAA) are available on Enterprise plans. Contact enterprise@oltodiscovery.com to discuss compliance requirements.

Data Residency

By default, data is stored in US-West-2 (Oregon). EU data residency and private cloud deployment are available on Enterprise plans.

Responsible Disclosure

If you discover a security vulnerability, please report it to security@oltodiscovery.com. We respond to all reports within 48 hours and maintain a responsible disclosure policy.

The Olto Discovery REST API is available for programmatic access to protocol generation, library management, and team resources. API access requires authentication via Supabase JWT tokens obtained through the standard auth flow.

Base URL

Authentication

Include the Supabase session token in the Authorization header:

Authorization: Bearer YOUR_JWT_TOKEN
Content-Type: application/json

Endpoints

Generate Protocol Request Body

{
  "researchGoal": "Investigate whether KRAS G12C knockout...",
  "field": "Oncology",
  "experimentType": "Gene editing (CRISPR)",
  "equipment": ["Flow cytometer", "PCR machine", "Western blot"],
  "budget": "Moderate ($10k–$50k)",
  "timeframe": "1–2 months"
}

Generate Protocol Response

{
  "protocol": {
    "id": "uuid",
    "title": "CRISPR-Cas9 Knockout of KRAS G12C...",
    "field": "Oncology",
    "score_feasibility": 91,
    "score_controls": 88,
    "score_reproducibility": 85,
    "score_clarity": 94,
    "sections": {
      "objective": "...",
      "hypothesis": "...",
      "design": "...",
      "materials": [...],
      "procedure": [...],
      "variables": {...},
      "controls": "...",
      "expected": "...",
      "quality": "...",
      "safety": "..."
    },
    "created_at": "2025-05-08T..."
  }
}