Guava
Guava provides a conversational voice AI platform delivered as a Python SDK and API, enabling developers to build deterministic, auditable voice agents for regulated industries. The platform combines proprietary ASR, TTS, and language models with structured code-based workflows instead of prompts, and includes built-in telephony, compliance certifications, and EHR integration capabilities.
- AI Agents
- Artificial Intelligence
- Developer Tools
- Software Only
- Healthcare Technology
Funding
Raised to date
$12.5MRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
across 1 round
CVWV$12.5M · Seed
Announced Nov 2026
Disclosed cumulative: $12.5M
Founders
Founder details are not available yet.
Product
Problem
Traditional voice AI solutions rely on fragile, prompt-based systems that accumulate conflicting rules and drift over time, making them impossible to maintain and trust for regulated calls. Companies building voice features face a massive engineering lift for telephony, latency, and compliance certifications, which delays product launches and stalls enterprise deals.
Solution
Guava provides a fully integrated voice agent API with a proprietary voice stack—ASR, TTS, and language models built together—shipped as a Python SDK. Instead of prompts, developers define call behavior through structured checklists and code callbacks, ensuring deterministic, auditable outcomes. The platform handles WebRTC and SIP telephony, turn-taking, barge-in, and carrier routing out of the box. It embeds as a white-label voice layer under a company's own brand, while the client's backend, model, and data remain the source of truth. Guava is pre-certified for SOC 2 Type II, HITRUST i1, and PCI DSS Level 1, with a production deployment ready in 14 days.
Target Audience
Primary customers are AI platforms, business process outsourcers, and healthcare organizations that need to embed compliant, reliable voice agents into their existing products or services without building the underlying voice infrastructure.
Features
- Deterministic code-based workflow engine where call behavior is defined by structured checklists and Python callbacks, not prompts, eliminating prompt drift
- Proprietary voice stack with ASR, TTS, and language models built together, reducing tail latency and failure modes compared to third-party API stitching
- Multi-channel telephony support including WebRTC for browser and SIP for PSTN on the same latency budget
- Pre-built compliance suite: SOC 2 Type II, HITRUST i1, PCI DSS Level 1, with BAA available on day one for healthcare
- 65+ fine-tuned workflows across regulated verticals, trained on 10B+ live agent minutes
- Pluggable RAG and intent recognition architecture, allowing teams to bring their own systems
- Agentic tenacity features: cross-channel continuity, context compaction, strategic re-contact, and zero-drift execution