# SixHelix — Full Content

> AI-native custom software development for US small and mid-sized businesses, priced per deliverable.

---

## Homepage (https://sixhelix.com/)

### Hero

**Build Anything — Output-Based Pricing**

Your budgets balloon and deadlines drift because token-based AI billing charges you for effort. We price by the deliverable, assuring you clarity on budget and timeline.

We price by the deliverable, assuring you clarity on budget and timeline.

### How It Works

**From Prompt to Product**

No discovery calls before you see a number. Price and scope are clear from the first sentence you type.

1. **Describe what you need** — Write a sentence or paste a spec. Our composer breaks it into modules, regions, and the exact deliverables your product requires.
2. **See a fixed, itemized quote** — Every module gets a price and a date — no token meters, no surprise overages. Toggle scope and watch the budget and timeline update instantly.
3. **Approve, and we ship** — You pay per deliverable. We build, you review at each milestone, and you only release payment when an output is accepted.

### The Difference: Token-Based vs Output-Based

**Token-based AI (Cursor, Lovable, Replit) — Pay for the meter:**
- Billed per token, per call — costs scale with effort, not value
- Estimates drift as prompts retry and context grows
- No fixed delivery date; "it's done when it's done"
- You absorb the risk of rework and dead ends

**SixHelix — Pay for the result:**
- Priced per deliverable — you know the total before we start
- Scope and budget locked the moment you approve
- Every module ships with a committed date
- You pay only when an output is accepted

### FAQ

**How is output-based pricing different from hourly or token billing?**
You pay a fixed price per deliverable, agreed before work begins. We don't bill for hours, retries, or AI tokens — the number you approve is the number you pay.

**What counts as a deliverable?**
Anything we can define and demo: a module, an integration, an agent, a screen. The composer breaks your project into these units so scope and cost stay transparent.

**What if the requirements change mid-project?**
You adjust scope and instantly see how budget and timeline shift. New deliverables are quoted the same way — no open-ended change orders.

**When do I actually pay?**
Per accepted output. You review each milestone and release payment only when a deliverable meets the agreed spec.

**Which regions and stacks do you support?**
We serve US small and mid-sized businesses on modern web and AI stacks — React, Node, Python, and LLM-integrated workflows. Our team operates across time zones to keep your project moving.

---

## Experts Page (https://sixhelix.com/experts)

**For Experts — Pick a task. Ship it. Get paid.**

Browse open work across the stack. Each task ships with a written spec, a test suite, and a fixed bounty — you only get paid when the output is accepted.

Experienced software engineers and AI builders can join SixHelix to take on well-specified bounty tasks. Each task has a defined spec, acceptance criteria, and fixed bounty. No ambiguous briefs, no scope creep.

Supported stacks: React, Node.js, Python, and LLM-integrated workflows (Claude, OpenAI, Anthropic APIs).

### How the bounty model works

- Every open task ships with a written spec, a test suite, and a fixed bounty amount.
- Pick a task, build to spec, submit for review.
- Payment releases only when the output is accepted against the acceptance criteria.
- No ambiguous briefs, no scope creep, no hourly negotiation.

### Open task categories

#### Web Development
Employee, candidate and admin surfaces — portals, dashboards and flows.

- **Core HR self-service portal** ($2,400) — Employee master data, org structure and a self-service portal for everyday HR actions. 18 specs, 59 test cases.
- **Leave management UI** ($1,200) — Configurable leave types, accrual policies and multi-level approval workflows. 12 specs, 39 test cases.
- **Expense & claim workflow** ($1,300) — Expense claims with receipt capture, policy checks and payroll reimbursement. 13 specs, 41 test cases.
- **Performance & OKR module** ($1,600) — Goal-setting, continuous feedback and review cycles tied to the employee record. 14 specs, 36 test cases.
- **Applicant tracking pipeline** ($1,900) — Applicant tracking system managing requisitions, stages and hiring decisions. 14 specs, 42 test cases.
- **Talent CRM datastore UI** ($1,800) — Searchable candidate datastore with consent tracking and re-engagement segments. 13 specs, 39 test cases.
- **Interactive org chart** ($1,100) — Drag-and-drop organisation chart with reporting lines and headcount roll-ups. 9 specs, 24 test cases.
- **Payslip viewer** ($900) — Employee-facing payslip view with breakdown, history and downloadable PDFs. 8 specs, 22 test cases.
- **Admin analytics dashboard** ($1,500) — HR ops dashboard surfacing headcount, attrition, leave and payroll cost trends. 11 specs, 28 test cases.
- **Onboarding wizard** ($1,400) — Guided multi-step onboarding flow with document collection and e-signatures. 10 specs, 26 test cases.
- **Approvals inbox** ($1,200) — Unified inbox aggregating leave, claim and request approvals for managers. 9 specs, 23 test cases.
- **Interview scheduling UI** ($1,500) — Calendar-driven booking surface with availability, reschedule and reminders. 10 specs, 33 test cases.

#### Agent Skills
Autonomous agents that source, screen, schedule and interview.

- **Consultant intake agent** ($2,400) — Client-facing agent that captures role requirements and shapes the hiring brief. 12 specs, 40 test cases.
- **Headhunter sourcing agent** ($2,200) — Sourcing agent that finds and engages candidates across marketplaces and the CRM. 13 specs, 43 test cases.
- **Screening & ranking agent** ($1,800) — Conversational screening agent that qualifies, scores and ranks applicants. 14 specs, 46 test cases.
- **Scheduling agent** ($1,500) — Scheduling agent that books interviews, syncs calendars and chases no-shows. 10 specs, 33 test cases.
- **Interviewer agent** ($2,600) — Interviewer agent that runs structured interviews and produces scorecards. 15 specs, 49 test cases.
- **In-app AI assistant** ($1,500) — Embedded copilot that answers HR & payroll questions and triggers self-service actions. 11 specs, 38 test cases.
- **Agent orchestrator** ($2,800) — Coordination layer that routes work between sourcing, screening and scheduling agents. 16 specs, 52 test cases.
- **Response evaluator** ($1,700) — Quality agent that grades agent outputs against a rubric and flags hallucinations. 12 specs, 34 test cases.
- **Conversation summariser** ($1,300) — Agent that distils long candidate threads into concise, structured briefs. 9 specs, 25 test cases.
- **Voice screening agent** ($2,000) — Outbound voice agent that conducts first-round phone screens with transcription. 12 specs, 43 test cases.
- **Safety & guardrail layer** ($1,900) — Policy layer enforcing PII redaction, tone and compliance across all agents. 13 specs, 37 test cases.

#### Integration
Two-way connectors into accounting, statutory boards and benefits.

- **Accounting sync (Xero · QB · SAP)** ($2,600) — Two-way sync that posts payroll journals and cost centres straight into the ledger. 14 specs, 47 test cases.
- **Statutory board e-filing** ($3,400) — Direct e-filing and contribution submission to each market's statutory authorities (CPF/IRAS/EPF/BPJS/SSS). 22 specs, 64 test cases.
- **Benefits provider connectors** ($1,800) — Connectors that reconcile enrolments and deductions with benefits providers. 10 specs, 33 test cases.
- **UCaaS connectors** ($1,050) — Push approvals and notifications into Slack, Teams and email inboxes. 8 specs, 25 test cases.
- **Messaging channels** ($1,200) — Conversational HR on Telegram, WhatsApp Business, and Messenger. 9 specs, 29 test cases.
- **Meta Business Messaging** ($2,400) — Unified Meta messaging across WhatsApp, Messenger and Instagram for candidates. 14 specs, 50 test cases.
- **Calendar sync (Google · Outlook)** ($1,200) — Bidirectional calendar integration for interview booking and availability. 9 specs, 28 test cases.
- **SMS gateway** ($900) — Two-way SMS for reminders, nudges and fallback when chat is unavailable. 7 specs, 25 test cases.
- **Bank disbursement files** ($1,700) — Generate and transmit bank-ready disbursement files for multi-country payroll. 12 specs, 34 test cases.
- **Telephony / voice provider** ($2,000) — Inbound and outbound voice with IVR routing and live transcription. 12 specs, 43 test cases.

#### Authentication
Identity, access control, SSO and audit across every tenant.

- **Role-based access control** ($1,500) — Role-based permissions and a full audit trail across the platform. 13 specs, 40 test cases.
- **SAML / OIDC single sign-on** ($1,700) — Enterprise SSO supporting SAML 2.0 and OpenID Connect providers. 12 specs, 38 test cases.
- **Audit log & session policies** ($1,300) — Immutable audit logging with configurable session and security policies. 11 specs, 30 test cases.
- **Multi-factor authentication** ($1,400) — TOTP, WebAuthn and SMS second factors with enforced step-up flows. 10 specs, 32 test cases.
- **SCIM user provisioning** ($1,500) — Automated user lifecycle via SCIM 2.0 with directory sync. 11 specs, 29 test cases.
- **Multi-tenant isolation** ($1,800) — Tenant scoping and data isolation guarantees across the shared platform. 12 specs, 35 test cases.
- **Consent & retention management** ($1,300) — Candidate consent capture with configurable data-retention rules. 10 specs, 27 test cases.
- **API key & token issuance** ($1,200) — Self-serve API key issuance with scoping, rotation and usage metering. 9 specs, 24 test cases.

#### Backend
Payroll engines, statutory calculations and the disbursement pipeline.

- **Indonesia payroll engine** ($2,200) — BPJS Kesehatan & Ketenagakerjaan, PPh 21 progressive tax & PTKP, THR & e-Bupot. 16 specs, 48 test cases.
- **Singapore payroll engine** ($2,000) — CPF (OW/AW) & age bands, SDL/CDAC/MBMF funds, IRAS AIS (IR8A) submission. 15 specs, 46 test cases.
- **Thailand payroll engine** ($2,100) — Social Security Fund (SSF), PND 1 withholding tax, provident fund contributions. 15 specs, 45 test cases.
- **Vietnam payroll engine** ($2,100) — Social/health/unemployment insurance, PIT brackets & reliefs, region rules. 15 specs, 45 test cases.
- **Malaysia payroll engine** ($2,300) — EPF (KWSP) & SOCSO (PERKESO), EIS contributions, PCB (MTD) & EA form. 16 specs, 47 test cases.
- **Philippines payroll engine** ($2,400) — SSS/PhilHealth/Pag-IBIG, BIR withholding (1601-C), 13th-month pay & alphalist. 17 specs, 50 test cases.
- **Gross-to-net calculation core** ($2,200) — Country-agnostic engine computing gross, deductions and net pay per cycle. 14 specs, 42 test cases.
- **GL posting service** ($1,800) — Maps payroll results to journal entries and posts to the ledger. 12 specs, 33 test cases.
- **Disbursement orchestrator** ($1,900) — Pipeline that batches net pay, generates bank files and confirms settlement. 13 specs, 36 test cases.
- **Talent CRM datastore** ($1,800) — Searchable candidate datastore with consent tracking and re-engagement segments. 13 specs, 39 test cases.

#### MCP & Harnesses
Model Context Protocol servers and the test harnesses that validate them.

- **Indeed MCP server** ($1,400) — MCP server that lets the agent search, post and apply on Indeed. 9 specs, 30 test cases.
- **LinkedIn MCP server** ($1,600) — MCP connector for LinkedIn sourcing, outreach and profile enrichment. 10 specs, 33 test cases.
- **Publishable custom MCP** ($2,200) — Publishable MCP surface for external platforms to invoke your agents. 13 specs, 42 test cases.
- **Agent evaluation harness** ($1,800) — Reproducible harness that runs agents against fixtures and scores outputs. 12 specs, 34 test cases.
- **Tool schema registry** ($1,500) — Central registry that versions and validates MCP tool schemas across servers. 10 specs, 28 test cases.
- **Sandboxed execution runtime** ($2,000) — Isolated runtime for executing MCP tool calls with resource limits and timeouts. 13 specs, 37 test cases.
- **Conversation replay harness** ($1,600) — Replays recorded conversations to detect regressions in agent behaviour. 11 specs, 31 test cases.
- **Usage metering & billing hooks** ($1,300) — Per-tenant metering of MCP invocations with billing event emission. 9 specs, 24 test cases.
- **MCP load & soak harness** ($1,500) — Performance harness that stresses MCP servers under concurrent agent load. 10 specs, 27 test cases.
- **MCP auth & key broker** ($1,500) — Auth broker issuing scoped, rotating credentials for MCP subscribers. 11 specs, 30 test cases.

---

## Blog: What Is SixHelix? The Six Strands Behind Every Software Build (https://sixhelix.com/blog/what-is-sixhelix)

*Published: June 12, 2026*

**TL;DR**

- SixHelix is an AI-native custom software service for US small and mid-sized businesses, priced per deliverable instead of per hour or per token.
- The name comes from the six strands every working software product is woven from: App Development, Agent Skills, Integration, Authentication, Backend, and MCP & Harnesses.
- You describe a project, see a fixed itemized quote with committed dates, and pay only when you accept each output.

### A story you have probably lived

Every business that has tried to buy custom software knows the shape of this story. It starts with a clear idea and a reasonable budget. Somewhere in the middle, the meter starts running. The budget balloons, the deadline drifts, and when you ask why, the answer is some version of "it took more effort than expected."

AI was supposed to fix this. Instead, token-based tools rebuilt the same meter in a new costume: you pay per token, per call, per retry. Costs scale with effort, not value. Estimates drift as prompts retry and context grows. There is no fixed delivery date — it's done when it's done — and you absorb the risk of every dead end.

We started SixHelix because we believe the problem was never the AI. It's the meter. Software should be bought the way every other business outcome is bought: a defined output, a fixed price, a committed date — and payment only when you accept the result.

### Why "SixHelix"? The six strands of a working product

The most famous structure in biology is the double helix: two strands wound around each other, each one useless on its own, together carrying the entire blueprint of a living thing. When we mapped the projects we'd shipped, the same pattern kept appearing — except the count was six, not two. Every piece of software that survives contact with a real business is six disciplines wound together. Leave one strand out and the structure unravels: the demo works, but the product doesn't.

1. **Helix 1 — App Development: the surfaces people touch.** Portals, dashboards, onboarding wizards, approval inboxes — the screens where your customers and your team actually live. This is the strand everyone pictures when they say "build me an app," and it's usually the only one a token-based tool quotes you for.
2. **Helix 2 — Agent Skills: software that does the work.** AI agents that don't just display information but act on it — intake, sourcing, screening, scheduling, interviewing — plus the orchestrators, evaluators, and guardrail layers that make agents safe to run. An agent without an evaluator is a liability with an API key.
3. **Helix 3 — Integration: meeting your stack where it lives.** Two-way connectors into the ledger (Xero, QuickBooks, SAP), calendars (Google, Outlook), Slack and Teams, WhatsApp and SMS, bank disbursement files, and statutory filings. Integration is what turns "a tool we bought" into "the way we work."
4. **Helix 4 — Authentication: the strand nobody asks for.** Role-based access control, SSO, multi-factor authentication, SCIM provisioning, tenant isolation, audit trails. Scoped on day one, because retrofitting trust is the most expensive deliverable of all.
5. **Helix 5 — Backend: the engines that must be exactly right.** Gross-to-net calculation engines, tax brackets and contribution rules, ledger postings, disbursement pipelines that move actual money. Where "mostly correct" means failure, and where written specs and test coverage matter more than demo polish.
6. **Helix 6 — MCP & Harnesses: the AI-native strand.** Model Context Protocol (MCP) servers let agents act on outside platforms; harnesses prove the behavior — evaluation rigs, replay suites, sandboxed runtimes, load harnesses. Helix 6 is how the other five strands get verified instead of trusted.

### Six strands, one fixed quote

A real project is never just one strand. A "simple HR portal" is a web surface (Helix 1) over a leave-accrual engine (Helix 5), synced to payroll and calendars (Helix 3), behind role-based access (Helix 4), with an assistant that answers policy questions (Helix 2), all proven by a test harness (Helix 6). You type a sentence — or paste a whole spec — into the composer, and it decomposes the project across the six strands into discrete, demoable deliverables. Every deliverable gets a fixed price and a committed date before any work starts.

### The comparison: paying for effort vs. paying for results

| | Token-based AI tools | SixHelix (output-based) |
|---|---|---|
| What you pay for | Effort — tokens, calls, retries | Results — accepted deliverables |
| Budget | Drifts as prompts retry and context grows | Locked the moment you approve the quote |
| Deadline | "It's done when it's done" | A committed date on every module |
| Risk of rework | You absorb it | We absorb it |
| When you pay | Continuously, as the meter runs | Only when you accept the output |

Token-based billing rewards inefficiency. Output-based pricing aligns our incentives with yours: ship the deliverable, get paid.

### FAQ

**What does the name "SixHelix" mean?** It refers to the six strands every working software product is woven from: App Development, Agent Skills, Integration, Authentication, Backend, and MCP & Harnesses. Like the strands of a helix, each is useless alone — a real product is all six wound together.

**Is SixHelix an agency, a dev shop, or an AI tool?** None of the usual categories. SixHelix is an AI-native custom software service: you describe a project, our composer breaks it into fixed-price deliverables across the six strands, and you pay per accepted output — no hourly rates, no retainers, no token fees.

**How does output-based pricing work?** Every deliverable gets a fixed price and a committed date before work starts. You review each milestone and release payment only when the output meets the agreed spec. Scope changes are re-quoted the same way — no open-ended change orders.

**Does every project need all six helixes?** No. The composer includes only the strands your project actually requires — but it quotes with all six in view, so nothing surfaces later as a surprise line item.

---

## Privacy Policy (https://sixhelix.com/privacy)

*Last updated: May 30, 2026*

### 1. Information we collect

We collect information you provide directly to us, such as when you describe a project using our prompt composer, contact us by email, or otherwise communicate with us. Types of information collected include:

- **Project descriptions and prompts** — text you enter into the composer to generate scope and pricing estimates.
- **Contact information** — your name and email address if you reach out to us directly.
- **Usage data** — pages visited, features used, referring URLs, browser type, device type, and approximate location derived from IP address.
- **Cookies and similar technologies** — small data files placed on your device to remember preferences and understand usage patterns.

We do not collect payment card numbers or other financial data directly. Any payment processing is handled by a PCI-compliant third-party processor.

### 2. How we use your information

We use the information we collect to provide, maintain, and improve our services; generate project scope, pricing estimates, and delivery timelines; respond to your comments, questions, and requests; send you technical notices, updates, security alerts, and support messages; monitor and analyze trends, usage, and activities on our platform; detect, investigate, and prevent fraudulent transactions and other illegal activities; and comply with legal obligations.

We do not sell your personal information to third parties. We do not use your project prompts to train AI models without your explicit consent.

### 3. AI processing and your prompts

SixHelix uses large language models (LLMs) to parse project descriptions, generate scope breakdowns, and calculate output-based pricing. When you type a prompt, it may be processed by one or more AI providers under data processing agreements that restrict secondary use.

We do not use your prompts to train third-party AI models. Prompts are processed solely to generate the response shown to you.

### 4. Sharing of information

We may share your personal information with third parties only in the following circumstances: service providers (cloud hosting, analytics, email delivery, AI inference) who are contractually restricted from using your data for any other purpose; legal compliance if required by law; business transfers in connection with a merger or acquisition; or with your consent for any other purpose.

### 5. Data retention

We retain personal information for as long as necessary to provide our services. Project prompts submitted without creating an account are retained for a maximum of 90 days for debugging and quality purposes, then deleted.

### 6. Security

We implement industry-standard technical and organizational measures including encryption in transit (TLS 1.2+), encrypted storage, access controls, and regular security reviews.

### 7. Your rights

Depending on your jurisdiction, you may have the right to access, correct, delete, restrict, or port your personal information, and to withdraw consent. Contact us to exercise these rights; we will respond within 30 days.

### 8. Cookies

We use cookies to collect usage data and remember preferences. We do not run advertising networks or sell cookie data to data brokers.

### 9. International transfers

Your information may be transferred to and processed in Singapore and the United States. We ensure appropriate safeguards are in place, including standard contractual clauses where required.

### 10. Changes to this policy

We may update this Privacy Policy from time to time. We will notify you of material changes by updating the date at the top of this page.

---

## Terms of Service (https://sixhelix.com/terms)

*Last updated: May 30, 2026*

### 1. Acceptance of terms

By accessing or using SixHelix, you agree to be bound by these Terms of Service. SixHelix is operated by SixHelix Pte. Ltd.

### 2. Description of service

SixHelix provides an output-based software development service. We accept project descriptions via our prompt composer, generate itemized scope and fixed pricing, and deliver agreed deliverables on committed timelines. A binding engagement is created only when both parties have signed a Statement of Work (SOW).

### 3. Output-based pricing model

All work is priced per deliverable as defined in an agreed SOW. Fixed price per deliverable — the price shown in an approved SOW will not change unless you request a scope change in writing. Payment for each deliverable is due upon your written acceptance of that output. Acceptance is deemed given if you do not raise written objections within 7 business days of delivery.

### 4. Intellectual property

Upon receipt of full payment for a deliverable, you own all intellectual property rights in that specific deliverable, including source code, documentation, and design assets. We retain ownership of any pre-existing tools, frameworks, and AI scaffolding used to produce deliverables, and grant you a perpetual, royalty-free license to use such components as embedded in the deliverable.

### 5. AI-generated content

Some deliverables may include code, copy, or designs that are wholly or partly generated by AI tools. We take responsibility for the quality, correctness, and fitness-for-purpose of all deliverables regardless of how they are produced.

### 6. Confidentiality

Both parties agree to keep confidential all non-public information received from the other party. This obligation survives termination for 3 years.

### 7. Warranties and disclaimers

We warrant that deliverables will conform to the acceptance criteria set out in the applicable SOW for a period of 30 days following acceptance. We will remedy defects reported during the Warranty Period at no additional charge.

### 8. Limitation of liability

Our total liability for any claims will not exceed the total fees paid under the applicable SOW in the 3 months preceding the claim.

### 9. Termination

Either party may terminate an SOW for convenience with 14 days' written notice. You will owe payment for all deliverables accepted before the termination date.

### 10. Governing law

These Terms are governed by the laws of Singapore. Disputes shall be resolved by arbitration administered by the Singapore International Arbitration Centre (SIAC).

---

## Contact

- Website: https://sixhelix.com
- Get started: https://sixhelix.com/#start
- Experts waitlist: https://sixhelix.com/experts