Trust, Operated
Continuously

SpiceFactory builds mission-critical software for partners in healthcare, automotive, and other regulated industries. Our security program is not a binder — it is an operating system: documented policies, named owners, evidence collected on cadence, and an append-only audit trail across the engagement lifecycle.

SOC 2, ISO 27001, HIPAA, FDA 21 CFR Part 11, and PCI DSS are the standards our control set is built to meet and exceed. The controls operate across every discipline of the studio — business development, design, engineering, quality, customer success, marketing, growth, and leadership. We make the live system available to partner security teams under appropriate access — not a point-in-time letter, the running record.

Studio ↔ Partner: Who Operates What

Most trust pages on the web belong to SaaS operators. SpiceFactory is not a SaaS operator — we are a product studio that engineers software the partner runs. That distinction shapes which controls live with us and which live with the partner, and we name it here on the public page so the responsibility split is never ambiguous to a CISO clearing the engagement.

Area SpiceFactory operates Partner operates
The studio's own environment Identity, endpoints, source-control hardening, secret management, dependency & static analysis, training, HR security, vendor management, BCP for our delivery infrastructure, this trust portal.
Engagement codebase & SDLC Secure-development practice, code review, change management, threat modelling, security regression, release sign-off, design-time data discipline. Acceptance, deployment, and change-approval governance under the partner's release cadence.
Partner production runtime Production infrastructure, key custody, production data, production uptime, production incident response, and the partner's own SOC 2 / ISO / HIPAA / FDA / PCI compliance obligations as the operator of the system.
Partner data in flight (during delivery) Data-classification handling, least-privilege scoped access, no-production-data-in-design rule, synthetic / anonymised data for development & test, data return or destruction at engagement close. Lawful basis for sharing the data, breach-notification obligations to data subjects, regulator-facing reporting.
Regulatory submission & certification Records-of-construction (audit trail, electronic-records discipline, IEC 62304-aligned SDLC, requirements traceability, evidence retention) to support the submission. Submission ownership and regulatory accountability as sponsor / operator. SpiceFactory does not sign as the regulated entity.
Subprocessors The studio's own subprocessors (productivity, source-control, secret stores, this portal's hosting) — disclosed publicly at /policies/subprocessors.html. Production-side subprocessors used by the partner's running system.

What the register covers, and what it does not. The 113 controls SpiceFactory maintains describe how the studio operates — how we engineer software, qualify engagements, handle partner materials, train people, and govern the program. It does not describe how the partner operates the system in production; that is the partner's posture, on the partner's trust page, against the partner's auditor. We engineer to support that posture; we do not stand in for it.

How We Protect Partner Data

SpiceFactory is an embedded product studio. The software we co-create sits inside our partners' security boundary, and we operate the controls below — continuously — to keep that boundary intact.

Access Control

Role- and engagement-scoped least-privilege access governed by our Access Control Policy. MFA enforced for all production and partner-data systems. Quarterly user access reviews are performed by the Compliance Officer with reviewer sign-off retained in the compliance register. Off-boarding triggers access revocation within one business day, evidenced by the offboarding checklist.

Encryption

Data in transit protected by TLS 1.2 or higher; data at rest protected by AES-256 on the platforms we control. Endpoint full-disk encryption (FileVault) is required on every team-issued device and verified via the quarterly endpoint-security attestation that every team member signs against the laptop baseline. Secrets stored only in dedicated secret managers — verified by GitHub secret scanning and push protection.

Secure Development Lifecycle

Documented in our Secure Development Policy and Code Review Standards. Branch protection requires independent peer review before merge; emergency hotfixes follow the documented exception process and are reviewed after the fact. SAST, dependency scanning, and secret scanning run on every pull request. Threat modeling is performed for features handling regulated data; OWASP ASVS guides security test coverage.

Incident Response

Operated under our Incident Response Plan with four severity tiers and documented response timelines. Partner notification follows the contractual SLA (default: within 72 hours of confirmation; HIPAA breaches per 45 CFR §164.410). Every Sev-2 and above triggers a post-incident review with root cause, corrective actions, and reviewer sign-off retained in the incident record.

Personnel Security

Pre-engagement verification — identity, right-to-work, references, and where lawful criminal record checks — performed for all personnel in accordance with the employment law of the jurisdictions we operate in (United States and Serbia). All personnel acknowledge the Information Security Policy and complete security awareness training at hire and annually; engineers on healthcare engagements complete HIPAA-specific training before access is provisioned. Onboarding and offboarding follow documented checklists with completion logged.

Vendor & Subprocessor Management

Vendors with access to partner data or production systems are assessed before engagement against our Vendor Management Policy. Critical vendors are reviewed annually — SOC 2 / ISO 27001 reports requested where available, contractual security and breach-notification obligations confirmed. The vendor register and review history are maintained in the compliance tracker.

Capabilities in Operation

The studio operates a defined set of security capabilities. The platforms below are the current implementations of those capabilities; the practice — what we do and how we evidence it — is the control, and the tooling rotates per engagement when a partner mandates a different stack. Configuration evidence is collected on a recurring cadence and retained in the compliance register.

Identity & access MFA enforced on the studio identity provider for every workforce account; admin and login audit logs retained at the studio-policy minimum; suspicious sign-in and admin-change alerts active. Currently operated on Google Workspace; the practice carries to any IdP a partner mandates.
Secret management Credentials and API keys held only in managed secret stores; no plaintext in source. Secret-scanning and push-time leak protection enabled on every engagement repository. Currently Google Secret Manager + GitHub Actions secrets + GitHub Advanced Security; substitutable per engagement.
Source-control hardening Branch protection on every default branch with required reviews and required status checks; force-push and default-branch deletion disabled; pull-request audit trail. Currently operated on GitHub; the practice applies to GitLab, Bitbucket, Azure DevOps, or any partner-mandated source-control host.
Dependency & static analysis Automated dependency vulnerability monitoring across every ecosystem we engineer in (npm, pip, Maven / Gradle, Cargo, Go modules, CocoaPods, Swift PM, Android Gradle, etc.); static application security testing in every CI pipeline with the analyser chosen to fit the language stack. Currently Dependabot, CodeQL, and Semgrep in active use; per-engagement equivalents for stacks they don't cover.
Endpoint posture Apple devices across the studio (laptops + iPhones) run against a documented laptop baseline — FileVault, current OS, screen lock, vetted-app posture (JetBrains WebStorm, Visual Studio Code, Claude CLI, OpenAI Codex CLI on the allow-list; anything else through app vetting), Apple Find My with iCloud-backed remote-wipe capability. Personal Android phones used for work enroll in the Google Workspace mobile-device policy on first sign-in, which provisions a Work Profile (business data separated from personal, screen-lock and encryption enforced, business-profile remote wipe available). Evidence layer for both: a signed endpoint-security attestation per team member (onboarding + quarterly) reconciled in a manual baseline review. Centralized third-party MDM is not deployed and is not required — the SOC 2 / ISO / HIPAA / PCI endpoint criteria are outcome-based, and Workspace's built-in mobile policy + the attested laptop baseline + identity-gated access meets them. Where a partner mandates a specific MDM at engagement level, we adopt it under their program.
Compliance register Internal register tracking owners, evidence, test cadence, last test date, next test date, and exceptions across every discipline of the studio. Changes recorded on an append-only audit log retained for six years (HIPAA-aligned). The register is the source of truth; policies and customer-visible surfaces read off it.

How This System Is Operated

The portal you are reading is the public face of a live trust system. Behind it sits our internal compliance register — the runtime where controls are owned, tested on cadence, and evidenced. Partners get the running record, not a snapshot.

A register, not a binder

Every control has a named owner, a test cadence, a last-test date, a next-test date, and attached evidence. Items moving out of currency are flagged on the operating dashboard the same day. The register is the source of truth — policy documents reference it, not the other way around.

Append-only audit trail

Every status change, evidence upload, access review, vendor review, and incident entry is recorded on an immutable audit log with actor, action, and timestamp. The log is queryable by leadership and surfaced to partner reviewers under NDA.

Continuous cadence, not annual cram

Access reviews quarterly. Vendor reviews on a tiered cadence by criticality. Policy reviews at least annually and on significant change. Risk and exception reviews on their own schedules. Items overdue are visible to leadership the moment they slip — not at audit time.

Built to exceed, not to pass

The control set is mapped to SOC 2 (2017 TSC), ISO/IEC 27001:2022 Annex A, the HIPAA Security Rule, FDA 21 CFR Part 11, and PCI DSS v4.0 (service-provider posture). We have chosen to operate the controls continuously rather than purchase a point-in-time attestation. Partner security teams are welcome to evaluate the live system directly.

“Compliance, run as a paperwork exercise once a year, doesn't make a studio safer — it makes a binder. We chose to operate the controls as a live system: every one of them has a named owner, a cadence, evidence on a clock, and a place on the audit log. Partner security teams who work with us get the running record — that is what trust looks like at SpiceFactory.”
Aleksandar Krtinić · General Manager
Information Security Officer · HIPAA Security Officer (45 CFR §164.308(a)(2)) · accountable owner of the SpiceFactory trust program
security@spicefactory.co

Whole-Studio Scope

Security at SpiceFactory is a whole-studio discipline — not an engineering function operating in parallel to everything else. Every discipline carries register entries, named owners, and a cadence. The same control posture answers a partner CISO whether the question is about engineering, design, customer success, marketing, or how the engagement was qualified at intake.

Business development & intake

Pre-engagement risk classification (Standard / Health / Payments / Medical-device / Safety-critical / Sanctions) before partner data is shared. Partner due-diligence: entity verification, sanctions screening, regulatory-posture review. Master engagement contracts (MSA, SOW, NDA, BAA, DPA, reliance addenda) reviewed before signature.

Design & UX

Engagement-scoped design-tool access; partner brand and asset confidentiality; no real production data in design artefacts under any circumstance; accessibility as a shipping quality gate; privacy- and security-by-design checkpoints at every design phase.

Engineering

Capability-named, technology-agnostic — the practice applies across every stack we engineer in (Node, TypeScript, Python, Scala / Java, Android, iOS, embedded, hardware, ML). MFA on every identity, branch protection on every default branch, automated dependency vulnerability monitoring, static analysis, secret-leak protection, change management, threat modelling.

Quality

Test data discipline — production data never enters non-production tiers; representative-data patterns and partner-supplied anonymised extracts only. Defect-to-security-incident escalation triggers. Three-signature release sign-off (engagement lead + QA + VP Eng). Security regression automation per engagement.

Customer success

Documented post-signature handoff. Engagement-health cadence (monthly + quarterly, scaled by risk class). Named escalation chain with audit-log entries on every escalation. Partner-satisfaction survey at midpoint and close. Post-engagement closeout checklist (data return, destruction, access revocation) within 30 days.

Marketing

Every external piece — case study, blog, talk, social, demo — reviewed before publication for partner confidentiality, regulated-data exposure, framework-claim accuracy. Partner-reference subregister with written approval per public mention. Marketing-data privacy under the same lawful-basis discipline we ask partners for.

Growth

Minimum-viable, privacy-preserving telemetry on studio properties — cookieless first-party by default; third-party trackers behind a consent gate. A/B experimentation ethics: accessibility preserved across variations, allocation never uses personal data. Mar-tech access under quarterly review.

Leadership

Standing ISMS oversight body (CEO, GM, VP Eng, COO) with a quarterly meeting cadence, written record into the audit log, and a quarterly signed attestation snapshot. Annual BCP and incident-response tabletops with full executive participation. Annual board-level security and compliance review.

Roles & Accountabilities

Security responsibilities are assigned to named roles, approved by executive leadership, and documented in the Information Security Policy.

Leadership

CEO

Executive Sponsor / CISO function

Approves the Information Security Policy and security budget. Owns the risk register at the executive level and accepts residual risk on behalf of the company.

General Manager

Information Security Officer & HIPAA Security Officer (45 CFR §164.308(a)(2))

Owns the policy framework and the control register. Coordinates quarterly access reviews, annual policy reviews, vendor reviews, and partner security reviews. Maintains the evidence trail.

VP of Engineering

Secure SDLC owner

Owns the Secure Development Policy, Code Review Standards, and Vulnerability Management Procedure. Approves engineering exceptions and signs off on production change requests.

Chief Operating Officer

Operational compliance owner

Owns recurring compliance operations: control test execution, training completion, HR security, and physical/operational controls. Reports operational metrics monthly.

All Personnel

Engineers

Secure development & data handling

Acknowledge the Information Security and Secure Development policies at hire and annually. Complete OWASP-aligned security training; HIPAA training prior to any healthcare engagement assignment. Access is granted only on assignment and revoked on rotation.

Design & UX

Privacy-aware design

Trained in data minimization. Real partner or end-user PII / PHI is not permitted in design tools or prototypes. NDA executed and background check completed before access to partner materials.

Product & Project Managers

Engagement security governance

Scope and document access per engagement. Confirm security requirements at kickoff against the Engagement Security Addendum. Track engagement-level compliance tasks in the compliance register.

Sales, Marketing & Business Development

Partner data stewardship

Acknowledge the Acceptable Use and Data Classification policies. Handle prospect and partner data per the Data Classification Policy. No access to engineering systems or partner code; endpoint encryption (FileVault) and screen-lock required by the laptop baseline and confirmed through the signed quarterly endpoint-security attestation.

Policy Framework

A documented policy set covers information security, data protection, incident response, secure development, HR security, vendor management, and HIPAA / FDA 21 CFR Part 11 alignment. Each policy is approved by executive leadership, version-controlled in Git, reviewed at least annually or upon significant change, and acknowledged by personnel at hire and annually. Available to partners under appropriate access controls.

Core Security

Information Security · Access Control · Password & Auth · Encryption · Acceptable Use

Operations

Change Management · Incident Response · BCP/DR · Vulnerability Mgmt · Risk Mgmt

Governance

Data Classification · Vendor Mgmt · Asset Mgmt · HIPAA Security · SDLC

People & Privacy

Remote Work & BYOD · Security Training · Subprocessors · Privacy Policy

View Policy Index Request Access

Engagement Lifecycle Controls

Security controls applied across the partnership lifecycle, governed by the Engagement Security Addendum and the Onboarding / Offboarding Procedure.

1. Engagement Start

Mutual NDA and Engagement Security Addendum executed. BAA executed prior to any PHI access. Background checks confirmed for assigned personnel. Engagement-specific data classification, access scope, and security obligations documented in the kickoff record.

2. During Delivery

Access scoped to named personnel and the systems required for their role. Pull requests require independent review under the Code Review Standards. SAST and dependency scanning gate the pipeline. Production data is not used in development or test environments; synthetic or de-identified data is used in its place.

3. Team Rotation

Joiners complete engagement-specific onboarding before access is provisioned. Rotation-off triggers the offboarding checklist: access removed within one business day, evidenced by the completed checklist and a corresponding audit-log entry.

4. Engagement Close

Access revoked within one business day of close. Partner data returned or destroyed within 30 days per the Data Retention Policy and any contractual terms. Credentials and shared accounts rotated. A written confirmation of data disposition is provided on request and retained.

Framework Alignment

Our control set is mapped to recognized frameworks so partner auditors can evaluate our posture against standards they already use.

Our position on certification. We have not purchased a CPA SOC 2 report, an ISO 27001 certificate, or a HITRUST certification — and that is a deliberate choice. The frameworks below describe a minimum bar to pass at a point in time; we operate the underlying control set continuously, against a register that names owners, mandates cadence, and holds evidence. Partner security teams reviewing our system look at the running record, not a letter. The mappings below describe what we operate against, and the evidence is available under appropriate access.
Independent assurance — in selection. We are in active selection for an independent CPA firm to conduct a SOC 2 Type 2 examination, and for a registered ISO/IEC 27001:2022 certification body. When the firms are engaged, the auditor names, scoping period, and reporting windows will appear here. Partners with strong views on which firms their own auditors prefer to receive reports from are welcome to influence selection — reach out to security@spicefactory.co.
Framework Scope of alignment Status Documentation
Trust Services Criteria — Security, Availability, Confidentiality
AICPA 2017 TSC, revised Points of Focus 2022
Common Criteria (CC1–CC9, Security) operated against on every engagement, plus Availability (A1) for studio delivery and the trust portal itself and Confidentiality (C1) for partner IP and engagement materials. Processing Integrity is added per engagement when SpiceFactory operates a partner-facing transactional system; Privacy is narrow-scope and applies only to controller-role data we hold directly (team HR, partner contacts, trust-portal access). Operated against — not certified TSC control mapping, policy set, operating evidence in the register
ISO/IEC 27001:2022 Annex A controls (93 controls across organizational, people, physical, and technological themes). ISMS scope: SpiceFactory product studio operations and engagement delivery. Aligned — not certified Annex A control mapping, ISMS context and Statement of Applicability
HIPAA (45 CFR Parts 160, 162, 164) Business Associate obligations under the Security Rule, Breach Notification Rule, and applicable Privacy Rule provisions. BAA executed prior to PHI access. Operating program HIPAA Security Program, BAA template, Breach Notification Procedure
FDA 21 CFR Part 11 Controls supporting partner regulatory submissions: electronic records integrity, audit trails, e-signature requirements, and IEC 62304-aligned SDLC for software intended as a medical device. Supports partner submissions Electronic Records Policy, 21 CFR Part 11 control mapping, FDA Compliance Program
PCI DSS v4.0 Service-provider alignment for partners with cardholder-data scope. SpiceFactory does not store, process, or transmit cardholder data as a standing service; where engagements engineer payment-handling code or operate payment-touching systems on a partner's behalf, our secure-development, key-management, access-control, logging, vulnerability-management, vendor-management, and training controls are mapped to the relevant PCI requirements so partners can rely on us under their PCI program. Aligned — service-provider posture PCI DSS control mapping, secure-development policy set, engagement-scoped reliance addenda
Evidence is refreshed per the control test cadence (typically quarterly for operating controls, annually for design controls). Stale evidence is flagged in the compliance register and remediated before the next quarterly review. Controls operate across every discipline of the studio — business development, design, engineering, quality, customer success, marketing, growth, and leadership — not engineering alone.

Frequently Asked Questions

Common questions from partners and prospects about our security practices.

Yes. We execute a BAA prior to any engagement involving PHI access, using either our template or the partner's. Our HIPAA Security Program documents the administrative, physical, and technical safeguards (45 CFR §§164.308, 164.310, 164.312) that apply when we operate as a Business Associate. The General Manager serves as the designated HIPAA Security Officer.
No CPA-issued SOC 2 report has been published — by design. We operate the AICPA Trust Services Criteria control set continuously against a register that records owners, test cadence, last/next test dates, and evidence. Partner security teams can review the live system under NDA; that is a stronger signal than a point-in-time letter, and we are happy to walk through any control or its evidence on request.
No ISO/IEC 27001 certificate is currently issued. Our ISMS is built and operated against the 2022 revision: documented scope, context, leadership commitments, risk treatment, Statement of Applicability, and the 93 Annex A controls. Mappings and evidence are available for partner review.
Yes. We support partners building software intended for FDA submission. Our SDLC follows IEC 62304-aligned practices for software lifecycle processes and 21 CFR Part 11 requirements for electronic records and electronic signatures, including audit trails, requirements traceability, and design controls. Final regulatory responsibility rests with the partner as the submission sponsor.
Yes — as a secure-development and engagement-support service provider. SpiceFactory does not store, process, or transmit cardholder data as a standing service; where an engagement engineers payment-handling code or operates a payment-touching system on a partner's behalf, our secure-development, key-management, access-control, logging, vulnerability-management, vendor-management, and training controls are mapped to the relevant PCI v4.0 requirements (Req. 6, 7, 8, 10, 11, 12 in particular). We can co-sign a PCI control-reliance addendum that names which requirements we are being relied on to satisfy and the evidence cadence behind each.
The studio engineers across Node.js, TypeScript, Python, Scala / Java, Android, iOS, embedded firmware, hardware, sensors, and ML / AI models — whatever the engagement requires. Our security practice is technology-agnostic by design: the control is the practice (multi-factor authentication on every identity, dependency vulnerability monitoring on every repository, secure-development and change-management on every codebase), and the per-stack tooling is implementation evidence. A partner's mandated identity provider, source-control host, MDM, or analyser substitutes for the studio default at the engagement level.
No. Security is a whole-studio discipline. Business development carries pre-engagement risk classification and partner due-diligence. Design carries privacy- and security-by-design checkpoints and the rule that production data never enters design artefacts. Quality carries test-data discipline, defect-to-incident escalation, and release sign-off. Customer success carries the engagement-health cadence and the post-engagement closeout. Marketing carries content review and the customer-reference subregister. Growth carries privacy-preserving analytics. Leadership carries the ISMS oversight body, tabletop participation, and the annual board-level review. Every discipline carries register entries with named owners and cadence.
Access is engagement-scoped and least-privilege. Provisioning and de-provisioning follow the Onboarding/Offboarding Procedure: joiners are provisioned only after assignment confirmation and completed engagement-specific training; rotation-off triggers access removal within one business day, recorded on the offboarding checklist and the audit log. Quarterly user access reviews cover all systems in scope.
Personnel use Apple devices (laptops + iPhones) operated against the studio laptop baseline — FileVault on, OS current, screen lock at five minutes, vetted-app posture (JetBrains WebStorm, Visual Studio Code, Claude CLI, OpenAI Codex CLI on the allow-list), Apple Find My / iCloud-backed remote-wipe capability — verified through a signed endpoint-security attestation at onboarding and quarterly, reconciled in a manual baseline review. Personal Android phones used for work enroll in the Google Workspace mobile-device policy on first sign-in: the device gets a Work Profile that separates business data from personal, screen-lock and encryption are enforced, and the business profile is remotely wipeable independent of the personal side. We do not run a separate third-party MDM product: the studio is small, every engineer is trained, Workspace provides the mobile policy on Android, the attestation is the laptop evidence layer, and the SOC 2 / ISO / HIPAA / PCI endpoint criteria are outcome-based — they are met by the baseline + attestation + Workspace mobile policy + identity gating + audit log. If a partner mandates a specific MDM at engagement level, we adopt it under their program.
Yes. We maintain reusable responses mapped to common questionnaire frameworks (CAIQ, SIG, vendor-specific). Typical turnaround is a few business days. Contact security@spicefactory.co to start.
In most engagements we build software that will process PHI but do not directly access or store live PHI. Development and testing use synthetic or de-identified datasets per the Data Classification Policy. Where an engagement requires live PHI access, a BAA is executed first and the HIPAA Security Program safeguards apply.
We follow the Incident Response Plan with four severity tiers and documented response timelines. Partner notification follows the contractual SLA (default: within 72 hours of confirmation; sooner if specified in the agreement). HIPAA breach notification follows 45 CFR §164.410. A post-incident review is performed for every Sev-2-and-above incident with root cause, corrective actions, and reviewer sign-off retained in the incident record.
Engagement infrastructure is provisioned under the partner's account wherever possible. SpiceFactory-operated systems supporting delivery (source control, productivity, secret management, this trust portal's hosting) are hosted with established cloud providers in regions agreed at engagement kickoff. The studio's subprocessor list — service, role, region, data category, BAA status, engagement-class scope — is published at /policies/subprocessors.html and reviewed quarterly, with partner notice 30 days before material changes. Engagement-specific subprocessors used by the partner's own production system sit on the partner's trust posture, not the studio's.
Each policy is reviewed at least annually, and upon any significant change to the threat landscape, technology stack, regulatory environment, or organizational structure. Reviews are recorded with reviewer, date, and version in the document control table.

Working With Your Security Team

Embedded partnerships are how we work. Bring us your questionnaire, your control walkthrough, your pen-test scope, or your disclosed vulnerability — we will respond with the named owner and the live evidence, not a sales contact.

Contact security@spicefactory.co