What is the LLM Council?

The LLM Council is Heavy3's multi-agent AI architecture built on three pillars: (1) different SOTA models with unique training and biases, (2) different adversarial roles that attack problems from different angles, and (3) different sources of truth—GPT searches Bing, Gemini searches Google, Grok searches X. True verification requires diversity at every layer.

How can Heavy3 help me verify my decisions?

Heavy3's LLM Council reviews your critical decisions like a board of expert advisors. Upload your strategy document, contract, or code and get feedback on blind spots, weak arguments, and risks—before you commit. Built by a scientist who needed accuracy for high-stakes decisions.

Why use multiple AI models instead of just ChatGPT?

Research shows multi-model ensembles beat single models by 13.2% on benchmarks. Each AI has different training and blind spots. When GPT-5, Gemini 3.1, and Grok 4 all agree your argument is weak—fix it. When they disagree, you see the risk before you commit.

Heavy3 was built by a scientist and engineer who regularly makes high-stakes decisions requiring depth and accuracy. After years of manually cross-checking AI responses before committing to important decisions, he built Heavy3 to automate the workflow.

What can Heavy3 verify?

Heavy3 verifies: business strategy and roadmaps, contracts and legal documents, pitch decks and investor materials, code architecture decisions, OKRs and planning documents—any high-stakes decision where accuracy matters and blind spots are costly.

Deep Audit is Heavy3's adversarial agentic research engine—the only one that makes AI agents actively argue with each other. Each agent gets a cognitive role: the Analyst digs deep, the Skeptic challenges assumptions, and the Lateral Thinker explores angles you didn't consider. They run 6 chained rounds of web research with parallel tool use, then a Synthesizer cross-examines all findings and delivers a verdict with academic citations. It catches what a single AI would confidently get wrong.

Is my data secure and private?

Yes—your data is encrypted, isolated, and never used for AI training. All data is encrypted in transit (TLS 1.3) and at rest (AES-256). We access AI models through OpenRouter with all privacy protections enabled: training disabled, prompt logging disabled. We never sell your data or use it for advertising. Staff access is limited to debugging and support only. SOC 2 compliant infrastructure with no tracking cookies. Built by a scientist who understands you're sharing sensitive information.

We're a 25-person B2B SaaS company in Ho Chi Minh City: $1.2M ARR, growing about 60% a year, and 60% of our revenue comes from customers in Singapore and Australia. We plan to raise a Series A in 2027. Should we flip to a Singapore holding company now, or stay a Vietnamese company? Weigh fundraising, tax, Vietnam's rules on outbound investment and moving capital abroad, hiring, and the cost of running two entities. Give me the biggest risks on each side, the strongest argument against your recommendation, and a clear verdict.

## Synthesized Answer ### Verdict **Flip to a Singapore parent — but treat it as a formal, sequenced restructuring that you begin now and complete before you open the 2027 Series A, not as a quick incorporation.** Establish the Singapore holding company promptly, capture your Singapore/Australia revenue through it, and run the Vietnam outbound-investment and founder-tax steps through the official channels. Do **not** execute the share swap until Vietnamese and Singapore counsel have cleared the tax and foreign-exchange path in writing. The three advisors converge on the core conclusion for the same underlying reason: at $1.2M ARR with 60% of revenue already from Singapore and Australia, you are past the point where an informal "flip" is viable, but you are also at the point where Vietnam's 2026 rules make a *legal* flip genuinely workable for the first time. The recommendation is conditional, and it is conditional on three things the advisors all independently flag: 1. **Go through the formal outbound channel.** Vietnam's 2025 Investment Law (in force since 1 March 2026) and Decree 103/2026 (in force since 3 April 2026) now explicitly allow a documented, arm's-length share swap — but only if you complete the outbound-investment procedure *before* the swap. A nominal-value swap or a "set up Singapore, sort out Vietnam later" route leaves an undocumented funds trail that surfaces in due diligence and tax inspections years later. It is not a real option. 2. **Give Singapore real substance.** A pure holding shell forfeits Singapore's new-startup tax exemption (which is explicitly unavailable to investment-holding companies), loses defensible tax residency, and invites Vietnamese transfer-pricing adjustments. The Singapore entity should contract with your Singapore/Australian customers and run genuine management functions there. 3. **Resolve the founder tax question before the swap, not during fundraising.** The biggest single financial risk on the "flip" side is Vietnam's 20% personal income tax on the capital gain deemed on transferring your Vietnamese shares — triggered while you receive only illiquid Singapore shares in return. Your valuation almost certainly breaches the VND 7 billion (~US$270–280k) outbound-investment threshold, so this must be priced, ruled on, or structured around in advance. The "stay Vietnamese" path is defensible only if your most likely Series A lead is a Vietnam-domiciled fund that invests directly into Vietnamese companies. Otherwise, staying means the flip becomes a condition of the round, executed under a financing deadline with minimum leverage. --- ## Detailed Findings ### 1. Fundraising — strongly favors Singapore The government's own startup materials note that investors commonly require Vietnamese startups to set up a Singapore holding company before investing. Vietnam's enterprise law lacks well-tested mechanisms for standard VC instruments (complex liquidation preferences, anti-dilution, convertible notes), which makes a Vietnamese entity a structural barrier to institutional capital rather than a mere preference. The HCMC Venture Capital Fund (launched 17 April 2026 under Resolution 98, ~VND 500bn / ~US$20M total) is real, but at that size it can anchor a round, not lead a foreign-syndicated Series A. It also confers city-level corporate tax exemptions (up to ~5 years) that you may forfeit if the structure becomes foreign-owned — a point to confirm with advisers. Two less-obvious fundraising benefits of flipping, surfaced by the Lateral Thinker and worth weighing: a Singapore entity opens access to a pre-A round, SAFE notes, and Singapore venture debt as bridge financing; and it reduces enterprise-procurement friction, because Singapore and Australian buyers face fewer compliance and withholding-tax hurdles paying a Singapore-registered vendor than a Vietnamese one. The counterpoint: flipping improves your *access* to capital, not your *growth*. At 60% growth you reach roughly $1.9M ARR by late 2027, which is thin for a Series A in the current market. A flip will not close a growth gap. ### 2. Tax — roughly even, slightly favoring Singapore once profitable (but not a windfall) **Singapore side:** 17% headline corporate rate, generally no capital gains tax. New companies get the startup exemption (75% of the first S$100k and 50% of the next S$100k, for three years), plus a 50% tax rebate for YA2026 capped at S$40k. The catch is critical: the startup exemption is **not** available to investment-holding companies, which is precisely why the Singapore entity must do real business, not just hold shares. **Vietnam side:** The new corporate income tax law introduces 15% (revenue up to VND 3bn) and 17% (VND 3–50bn) rates based on the *previous year's* revenue. But VND 50bn is roughly US$1.9–2.1M — at your growth rate you cross that around 2027–28 and pay the standard 20% anyway. Moreover, advisers note the reduced rates generally do **not** apply to subsidiaries whose related party is above the threshold, and most foreign-owned subsidiaries don't qualify. Plan on 20% on a cost-plus service margin for the Vietnamese operating entity. **Two meaningful cross-border points:** Vietnam's dividend withholding to foreign parents is currently 0%, and profit repatriation timelines were relaxed (profits must return within 12 months of distribution, up from 6). Neither is a reason to stay Vietnamese, but both reduce the ongoing tax friction of a Singapore parent. **Transfer pricing is the permanent compliance burden.** With 60% of revenue landing in Singapore and 25 employees (your cost base) in HCMC, intra-group service fees and royalties must be arm's-length and documented from day one. Price the markup too low and Vietnam alleges evasion; too high and Singapore scrutinizes margins. ### 3. Vietnam's outbound-investment and capital-controls rules — the binding constraint This is where 2026 changed the old advice materially: - **New Investment Law (1 March 2026):** no more policy approval for outbound investment; fewer projects need a certificate; most only need foreign-exchange registration. - **Decree 103/2026 (3 April 2026):** projects under VND 7bn (~US$270–280k) outside restricted sectors can skip the outbound investment certificate; share swaps are explicitly accommodated if the outbound procedure is finished first and an arm's-length value is documented. - **Circular 34/2026 (31 July 2026):** replaced the outward-investment foreign-exchange circular; cash transfers must use a dedicated outbound-investment capital account at a licensed Vietnamese bank. - **Pre-registration remittance cap:** at most 5% of project capital and never more than US$300k — not a license to move operating cash offshore early. - **New foreign-owned rule:** a Vietnamese company that becomes more than 50% foreign-owned cannot invest abroad until it shows two consecutive profitable years. This means any future Australian subsidiary should sit under Singapore, not Vietnam. The Direct Expert's sharpest point bears repeating: a share swap does **not** bypass these rules, and the inbound half matters too — the Singapore parent acquiring control of the Vietnam company triggers foreign-investment registration and capital-account procedures in Vietnam. "Incorporate a parent, sign a nominal-value swap, and assume the ownership chain is finished" is the single most common and most damaging error in this structure. The valuation point cuts against delay: your Vietnamese shares will almost certainly be valued above VND 7bn, so expect to need the outbound investment certificate, and expect the founder tax bill to rise with every quarter of ARR growth. ### 4. Hiring — moderately favors Singapore, with one Vietnam-specific wrinkle Your 25-person team can remain employed by the Vietnam entity either way. The benefit of a Singapore parent is an employee stock option plan at the Singapore level, which is standard and legible to investors and senior regional hires. The wrinkle: Vietnamese staff receiving shares in a foreign parent go through a separate State Bank procedure, so the option plan must be designed with a Vietnam-specific implementation review from day one. You do **not** need to hire in Singapore — a locally resident director (a paid nominee is acceptable to start) plus a company secretary satisfies the requirement. ### 5. Cost of two entities — modest but real The Singapore government filing fees are trivial (S$315 to register, S$60 annual return). The real cost is professional services. The Lateral Thinker's estimates: a one-off flip cost of roughly **US$40–100k** (lawyers in both countries, an arm's-length valuation, and founder-tax advice), plus **US$20–40k a year** in recurring costs (Singapore director and secretary, accounting, transfer-pricing documentation, and dual audits). That is roughly 2–3% of ARR — justified if it improves your Series A odds even modestly, but it is a permanent overhead, not a one-time expense. --- ## Biggest Risks on Each Side ### If you flip (and get it wrong) 1. **The "phantom tax" (highest severity).** Vietnam's General Department of Taxation treats the share transfer as a taxable event. Individuals transferring capital in an LLC face 20% personal income tax on the net gain (and, where cost basis cannot be established for certain securities transfers, 2% of proceeds). The GDT can deem fair market value on a business with $1.2M ARR and 60% growth. You could owe a substantial cash tax bill while holding only illiquid Singapore shares. 2. **A defective transaction.** Missing outbound or inbound approvals, an unsupported valuation, an unworkable funds trail, or routing money through personal accounts instead of the dedicated investment capital account. These are foreign-exchange breaches that surface in tax inspections and kill due diligence years later. 3. **Transfer-pricing and tax-residency risk.** If the Singapore entity has no substance, you risk Vietnamese tax adjustments *and* loss of Singapore tax residency and its benefits. Singapore incorporation alone does not create Singapore tax residency. 4. **Losing Vietnam-only incentives** (HCMC's startup tax exemptions, if they don't apply to foreign-owned entities), and the two-consecutive-profitable-years lock on future outward investment from the Vietnamese entity. ### If you stay Vietnamese 1. **Series A execution risk (highest severity).** Foreign VCs are reluctant to lead directly into a Vietnamese entity. If a lead asks for a Singapore parent, you resolve valuation, tax, approvals, banking, contracts, and employee equity under a financing deadline — the worst possible negotiating position. Term sheets are not binding, so a promised investment conditioned on a flip may never arrive. 2. **The flip gets more expensive every quarter.** The arm's-length valuation rises with ARR, so the founder tax bill and the threshold breach both worsen as you wait. 3. **Exit illiquidity.** Acquirers prefer not to buy a Vietnamese entity directly, and moving capital out of Vietnam for distributions or vendor payments is subject to strict FX controls. 4. **Friction with SG/AU enterprise buyers** who prefer a locally registered vendor, and harder option plans for recruiting senior or regional staff. --- ## The Strongest Argument Against the Recommendation **You may pay for a complex restructuring and a permanent dual-entity overhead only to discover that your eventual lead would have invested directly in Vietnam — or that your ARR does not yet justify the cost.** Vietnam's rules changed faster in 2026 than in the prior five years combined, and officials now argue startups no longer need a Singapore headquarters to obtain tax benefits. Commentators have noted Decree 103's stricter conditions could shut out high-growth startups. On this view, spending US$50–100k and three-to-six months of founder time on a structure for a round your ARR may not yet support is premature — you could keep Vietnam's incentives, raise a Vietnamese pre-A, and flip only when a named lead insists and effectively pays for it. **Why the recommendation survives this argument:** the share-swap rules require the outbound procedure to be finished *before* the swap, so the work cannot be compressed into a term-sheet deadline; the swap is valued at arm's length, so it gets more expensive every quarter; the Vietnamese small-company tax rate disappears on its own within ~18 months; and the HCMC fund is too small to lead a US$5–10M round. The only scenario that flips the verdict is a specific, credible Vietnam-domiciled lead willing to underwrite the Vietnamese structure directly. --- ## Points of Disagreement **1. Timing of the share swap.** The Direct Expert and Lateral Thinker say complete the swap before you open the 2027 process. The Critical Analyst frames the ideal as *delay the swap until you have a term sheet* (to use incoming VC funds to offset restructuring costs), then immediately concedes this creates a "chicken-and-egg" trap where top-tier VCs pass on a messy unresolved structure. On balance, the pre-Series-A completion view prevails: the rules require sequencing the outbound procedure first, and doing so under a financing deadline is exactly the failure mode you are trying to avoid. **2. How much operating substance Singapore needs.** The Direct Expert warns against moving your team or existing revenue "on paper," while the Lateral Thinker argues Singapore *should* sign and bill your SG/AU customers to establish commercial substance. These are reconcilable — the shared principle is that Singapore must have genuine substance, and a paper flip solves nothing — but the degree of revenue migration is a genuine judgment call your counsel should structure deliberately, because moving IP out of Vietnam is itself a taxable event. **3. Whether the founder tax is manageable or catastrophic.** The Critical Analyst treats the 20% deemed capital-gains tax as the primary failure mode; the Direct Expert and Lateral Thinker treat it as material but manageable through an arm's-length valuation and transaction-specific structuring. The truth depends on facts none of us can verify from outside: your legal form (joint-stock company vs LLC), your founders' cost basis, and the valuation methodology. This is precisely why the decision gate below makes a binding tax ruling or documented structure a precondition. --- ## Decision Gate Commission one coordinated Vietnam–Singapore legal and tax memo **now**, covering: the cap table and each founder's tax position, an independent arm's-length swap valuation, the outbound (OIRC/foreign-exchange) and inbound (foreign-acquisition) approval path, bank-confirmed money flows, IP and customer-contract treatment, the employee equity plan's Vietnam procedures, and an itemized first-year and annual budget. In parallel, sound out your most likely 2027 lead investors. - **Proceed this year** if counsel can document a compliant, affordable route and credible leads prefer the Singapore parent. - **Stay Vietnamese-parented for now** if the tax cost or approval path is disproportionate — but only while actively confirming you have a Vietnam-based lead willing to invest directly. Do not present a paper flip as a completed fundraising solution. A Singapore incorporation certificate without the completed outbound procedure and a defensible tax position is a liability, not an asset, in due diligence. --- ## References [1] KPMG Vietnam – Decree 103 on outbound investment. https://kpmg.com/vn/en/insights/2026/07/decree-103-on-outbound-investment.html [2] Duane Morris – Vietnam: The New Investment Law 2026. https://blogs.duanemorris.com/vietnam/2025/12/18/vietnam-the-new-investment-law-2026-what-you-must-know/ [3] Vietnam Government – Law on Investment 2025. https://en.baochinhphu.vn/viet-nams-law-on-investment-2025-111260731145214387.htm [4] ALIAT Legal – Decree No. 103/2026/ND-CP: Simplification of Outbound Investment Procedures. https://www.aliatlegal.com/resources/legal-updates-en/decree-no-103-2026-nd-cp-simplification-of-outbound-investment-procedures/ [5] State Bank of Vietnam – Circular 34/2026 on outward-investment foreign exchange. https://asemconnectvietnam.gov.vn/law.aspx?ID1=2&MaVB_id=3164&ZID1=10 [6] Acclime Vietnam – Capital Gains on Disposal of Shares/Capital: A Quick Guide (2026). https://vietnam.acclime.com/tax-updates/capital-gains-on-disposal-of-shares-capital-a-quick-guide/ [7] PwC – Vietnam Corporate Income Determination (Worldwide Tax Summaries). https://taxsummaries.pwc.com/vietnam/corporate/income-determination [8] Vietnam Government – Guidance on personal tax on capital and securities transfers. https://baochinhphu.vn/thue-thu-nhap-ca-nhan-doi-voi-chuyen-nhuong-chung-khoan-10226032718480641.htm [9] BDO – Vietnam New Corporate Income Tax Law. https://www.bdo.global/en-gb/insights/tax/world-wide-tax/vietnam-new-corporate-income-tax-law-affects-nonresident-businesses [10] Sleek – Startup Tax Exemption Singapore 2026. https://sleek.com/sg/resources/startup-tax-exemption-singapore/ [11] PwC – Singapore Corporate Taxes on Income (Worldwide Tax Summaries). https://taxsummaries.pwc.com/singapore/corporate/taxes-on-corporate-income [12] Singapore ACRA – Director requirements and filing fees. https://www.acra.gov.sg/register/business/registering-different-business-structures/local-company/appointing-company-directors-other-key-officers/ [13] Vietnam Briefing – Profit Repatriation in Vietnam 2026. https://www.vietnam-briefing.com/news/profit-repatriation-vietnam-brief-guide-2026.html/ [14] Indochina Link – Profit Repatriation Vietnam: Remitting Dividends Abroad. https://indochinalink.com/profit-repatriation-remitting-dividends-abroad/ [15] Kindrik – A primer on how to flip into Singapore for SEA startups. https://kindrik.sg/guides/a-primer-on-how-to-flip-into-singapore-for-southeast-asian-startups/ [16] VietNamNet – A turning point for Vietnam's startup funding landscape. https://vietnamnet.vn/en/a-turning-point-for-vietnam-s-startup-funding-landscape-2509810.html [17] NSSC – Current State of Startup Investment in Vietnam 2025. https://nssc.gov.vn/startup-stories/insights/current-state-of-startup-investment-in-vietnam-2025/ [18] PitchBook – Southeast Asia Private Capital Breakdown. https://files.pitchbook.com/website/files/pdf/2024_Southeast_Asia_Private_Capital_Breakdown.pdf [19] Vinex – Vietnam Outbound Investment: New Lending Rules from August 2026. https://www.vinex.com.vn/post/vietnam-outbound-investment-steps-and-the-new-70-cap ### Confidence: 8/10 The three advisors are directionally aligned on the verdict and on the core legal facts (20% founder tax, VND 7bn threshold, Decree 103/Circular 34 effective dates), but Decree 103 is only ~6 months old and its guidance is still being filled in — and the founder-tax severity depends on your legal form and cost basis, which require transaction-specific counsel. Treat the verdict as high-confidence; treat the tax and approval path as needing written confirmation before you execute.