Investor note

Bazas — investor note

Pre-Seed · Last updated October 7, 2026

For informational purposes only. Nothing here constitutes an offer to sell or a solicitation to purchase securities.

Executive summary#

Service businesses pay for leads through Yelp, Thumbtack, Google, and other channels, then lose a share of them because no one answers fast enough. Bazas connects supported inbound conversations, agreed scope, appointments, jobs and customer documents in one operating context. AI assistance is introduced with human review and a workflow agreed with each pilot. The wedge is inbound; the expansion is the operating memory and workflow for the business.

Problem#

A customer calls or messages while the owner is busy. The response is delayed, the customer moves to the next provider, and the business loses both the lead cost and the job. The pain is concentrated in three places: missed calls, fragmented inboxes across channels, and inconsistent follow-up.

Product & status#

October 6 developer reports describe implemented customer conversations, records, appointments, jobs, estimates and the core customer portal. Estimates can keep the agreed version and customer-facing materials connected, with separately approved additions. Pilot prepares drafts; the operator reviews important actions and sending. Company branding retains Bazas attribution.

Selected invoice and manual-payment processes are reported implemented. A customer's transfer notice is not proof of money received. The chosen process must reconcile confirmed payments, balances, receipts and optional accounting allocation. QuickBooks is an optional integration, not a requirement for every pilot. An embedded Bazas card checkout is planned and is not offered as ready.

Supported communication channels, hours, calendar availability and transfer during business hours require configuration and acceptance. Browser calling and expanded after-hours handling are separate development releases. Internal iPhone trip recording requires consent; the driver's detailed journal remains private. Real distance and export acceptance are not yet established by these reports.

The product is in private beta with founder-led onboarding, and access to beta features is limited to selected pilot customers. See the capabilities and their pilot limitations.

Web application: The web workspace implements customer records, conversations, jobs, appointments, estimates, the core customer portal and selected invoice/payment processes. The current rollout is a closed, founder-led pilot: access, configuration and an agreed workflow are checked with each company before use. iOS application: The October 4 Apple snapshot showed build 91 in internal TestFlight. October 6 developer reports describe newer mobile work but do not verify the currently available Apple build or external access. The available build and its selected workflows need confirmation before an external pilot; full web parity and public App Store availability are not claimed.

Scope reviewed against developer reports dated October 6, 2026 and the earlier source review. These are implementation reports, not a new independent runtime audit. Onboarding confirms deployment, configuration and acceptance for each company's chosen process. Product evidence date: 2026-10-06.

A context-aware dispatcher, introduced in stages#

The October 7 founder-approved programme extends the existing CRM into a context-aware dispatcher. Product acceptance and commercial evidence remain separate from this implementation scope.

We are implementing this connected dispatcher experience. Start with an agreed pilot; expand automation as each workflow is checked with your company.

• Platform: Permissions and action limits
• Company: Your team and business rules
• Service: Scope, questions and pricing
• Source: The rules for this enquiry
• Request: The customer and what was agreed

Enquiry sources carry their own automation policy even when the conversation changes channel. Human-only work can retain AI suggestions without external effects. Mandatory approvals, individual pauses and operator intervention stay explicit.

• Ask what matters. Remember what was agreed. Use your services, pricing and the customer’s history. Collect only the missing essentials, keep promises visible and bring the next useful step to the team.
• A new channel. The same conversation. A marketplace enquiry can continue by SMS or email without losing its context. We are connecting source-specific rules, email threads and human handoff across supported channels.
• Plan the visit around the real work. Coordinate suitable technicians, availability, job duration and travel. Handle changes and cancellations, and alert a person when a decision or response is missing.
• Turn agreement into clear documents. Connect preliminary estimates, customer approval, attachments, additional work and the invoice. Keep the accepted scope visible instead of asking your team to enter it again.
• Follow the payment. Not just the message. A customer saying “paid” is not confirmed funds. Connect verified or manually recorded payments, balances and receipts; update reminders when the situation changes.
• Stay helpful after the job. Build a clear path for feedback, complaints and warranty questions. Keep service follow-up separate from marketing, and stop when there is no useful next action.

Acceptance before broader delegation

1. Suggestions without external actions
2. Human-confirmed execution
3. Limited agreed automation
4. An accepted end-to-end workflow

Quality and operating economics

• Recommendations accepted, edited or rejected
• Correct completed actions and delivery failures
• Useful response time and human effort
• Total cost per completed step, including setup and support

No new external customer results, measured savings or runtime acceptance are announced by this update.

Why now#

• Fast, useful responses may help businesses retain inbound inquiries.
• Businesses that already pay for leads have a practical reason to measure missed inquiries.
• Inbound demand is fragmented across calls, texts, email, and lead marketplaces.
• AI-assisted responses can reduce repetitive work when tested against a company's actual services and constraints.
• Modern APIs let voice, messaging, scheduling, and CRM connect in one workflow.
• Founder-assisted pilots can compare support effort and software cost with the work saved.

These are pilot hypotheses. No external market statistic or measured effect is claimed here.

Traction & methodology#

We do not publish unverified figures. Each metric we share carries a defined period, an as-of date, and a methodology (for example, “unique inbound inquiries processed in pilot workspaces”). Request the evidence currently available. Pilot counts, revenue, retention and ROI are not publicly verified here; free pilot access is not counted as paid revenue.

Business model#

A monthly subscription includes defined AI, voice, and messaging allowances. Revenue can expand through higher usage tiers, overages, additional users, and additional locations.

Current: Monthly subscription per business, Included AI, voice & SMS allowances, Usage overages beyond the plan allowance. Near-term: Higher usage tiers, Additional users, Additional locations, Premium workflows. Long-term: Additional financial automation, Agency & franchise plans, Company branding & agency workflows, Enterprise integrations. Investor-facing economics (blended ARPA, gross margin, retention) require dated evidence before disclosure; availability is confirmed per request.

Market#

Bazas starts with small U.S. service businesses, initially in New York City: an owner or a small team managing inquiries, quotes and work. Contractors and small agencies are additional pilot audiences with separately agreed scope.

Sizing is bottom-up: Target businesses × expected annual revenue per account = serviceable revenue opportunity. Numeric inputs and a sourced market estimate are not yet published; any later model must provide a source, date and confidence for each input. We do not claim the entire SMB or global AI market.

Go-to-market#

Phase 1 — Founder-led (Live); Phase 2 — Repeatable vertical motion (Planned); Phase 3 — Channel expansion (Planned). A relationship counts as a partnership only when it is live, not when there is potential interest.

Competition & defensibility#

Buyers can choose existing field-service software, a separate AI receptionist, a human office service or a stack of specialist tools. Capabilities vary by vendor, plan and configuration; this site does not score whole categories without evidence. The intended Bazas proposition: Available customer context follows the inquiry into agreed work and documents, AI prepares the next step with explicit human review and configured permissions, Customer-facing documents and the portal remain connected to the operating process, A founder-led pilot accepts the chosen workflow before adding automation. These advantages are being built and are not yet a proven long-term moat.

Roadmap & milestones#

Now: Closed web pilots with founder-led fit and setup; Connected inquiries, agreed documents, work and the core portal; Human-reviewed AI actions and confirmed channel permissions; Chosen internal iPhone scenarios; current build checked separately; Define the payment process, support contact and result to measure; Implement context-aware dispatcher suggestions without external actions. Next: An external company completes the agreed process; Accept dispatcher suggestions, human-confirmed execution and limited source-specific automation; Amounts, confirmed payments, receipts and accounting allocation reconcile; Real trip distances and a first accepted PDF/CSV export; Repeatable onboarding, role checks and issue tracking; Measure recommendation acceptance, useful response time and total cost per completed step; Accept phone recipients, after-hours handling and browser calls as separate releases. Later: Embedded card checkout after payment acceptance; Multi-company dispatch and broader contractor settlements; Expand accepted dispatcher workflows through confirmed payments, aftercare and warranty handling; Broader integrations and field workflows chosen from pilot demand; External mobile distribution and public App Store release after their gates; Self-service onboarding once setup and support are repeatable. Milestones are shown on the homepage roadmap. These are sequencing goals; numeric adoption, revenue and margin targets have not been published.

Team & corporate structure#

Bazas is founder-led today (Nikita Lylov, Founder & CEO). The operating company is Bazas AI LLC; the definitive issuing entity, ownership, and IP assignment are set out in the offering documents and the data room.

Fundraising & use of funds#

Bazas is raising a Pre-Seed round to prove reliable inbound workflows, measurable customer outcomes, a repeatable acquisition channel, and unit economics within target. Round terms are shared with qualified investors after securities-counsel review. Use-of-funds categories are shown on the homepage; exact allocation is shared with investors.

Risks#

• Product reliability. AI can misunderstand an inquiry, and connected channels or documents can fail to deliver the expected result. The selected process needs real acceptance and a visible recovery path.
• Customer adoption. Independent companies may not find enough everyday value, trust the AI or accept the setup effort. Usage in the founder's business does not establish external adoption.
• Competition. Mature FSM and AI-receptionist companies could add similar capabilities.
• Unit economics. Voice, SMS, AI, and support costs could compress gross margin.
• Platform dependency. Bazas depends on third-party APIs, telecom, messaging, and lead-source integrations.
• Execution risk. A Pre-Seed company has limited team and capital.
• Data & privacy. The system processes conversations and customer data.
• Fundraising risk. A future round may not close on the desired timeline or terms.

Mitigations reduce, but do not eliminate, these risks.

Founder letter#

I've worked with CRM systems since 2011 — in sales, as an entrepreneur, and running my own service business, Repair ASAP, in New York City. Bazas exists because I lived the missed-lead problem: paid leads arriving while the team was out on jobs, conversations fragmented across calls, SMS, email, Yelp, and Thumbtack, and customer context lost between first contact, booking, and follow-up. Existing tools managed parts of the work, while important context still needed to be carried between them.

Bazas was developed and tested inside that live operating workflow. Repair ASAP — my own service business — is Bazas's first operating customer and real-world test environment, so results to date are internal operating validation rather than independent third-party proof. The hardest part has been reliability and trust: owners rightly want control over what an AI says to their customers, which shaped the product — owner-defined rules, human takeover, and safe defaults first. What we must prove next is that this converts inbound into booked work repeatably, at healthy unit economics, across service businesses beyond a single operator.

Founder: Nikita Lylov, Founder & CEO — New York City · LinkedIn.

Last updated & changes#

• October 7, 2026 — Customer portal, document approvals and payment limits clarified from October 6 developer reports; alternatives replace unsourced category scores. The outcome-led delivery sequence now includes all eighteen priorities. This editorial update does not represent a new runtime, Apple or external pilot verification.

This website is provided for informational purposes only. Nothing on this website constitutes an offer to sell, a solicitation of an offer to buy, or a recommendation regarding any security. Any offering, if made, will be conducted only through definitive offering documents and in accordance with applicable law.

Request full materials