METLIVI PTE. LTD.
By METLIVI PTE. LTD.We built Metlivi—a healing, social mobile world powered by emotional AI—and have productized the same underlying capabilities as Frugine (frugine.com), an enterprise-facing AI toolflow and operations console. Frugine covers the critical path from development to launch: member management, roles/permissions, prompt management (with versioning), testing sandboxes, product evaluation, task center, analytics, and bug tracking. On this foundation, partners can integrate our emotion engine, long-term-memory intelligent agents/NPCs, generative content services, and voice/multimodal APIs to rapidly build emotion-friendly, positive-psychology-oriented customized experiences (e.g., branded virtual worlds/interactive marketing, training and coaching simulations, Gen Z–facing customer engagement), and use analytics dashboards to quantify engagement, retention, and conversion. Frugine is currently being piloted with two established enterprises; our go-to-market focus is Southeast Asia, China, and Japan/Korea, with localization and compliant deployment supported and delivery possible within weeks. Frugine is currently in pilot with two established companies, with the initial rollout focused on Southeast Asia, China, and the Japan/Korea markets.
- Stage
- Prototype
- Industries
- Healthcare
- Media and entertainment
- Technology
- Deployment
- Cloud (SaaS)
Why it is worth it
Unlike general LLM platforms or emotion-AI point tools, Frugine combines a governed, end-to-end toolflow with emotion-aware, memory-based agents. In one console you manage prompts, run sandboxes, evaluate, orchestrate tasks, and analyze outcomes; drop-in APIs turn pilots into production. The stack is model-agnostic and RAG-ready, deployable to cloud VPC or on-prem, localized for SEA/China/Japan/Korea—so enterprises launch measurable, low-pressure experiences in 6–8 weeks, without vendor lock-in.
- Enterprises tell us three obstacles stop GenAI from creating business value:
- 1) Fragmented dev workflow. Disconnected tools slow iteration and create audit gaps. Frugine consolidates the pipeline inside one console—from prompt management & versioning, testing sandboxes, and product evaluation to task orchestration, analytics, and bug tracking—so engineering, compliance, and business teams work off the same source of truth.
- 2) Slow time-to-pilot / time-to-value. New experiences typically take months to assemble. Our reusable toolflows and templates standardize the path to a 6–8-week PoC, with governance built-in (RBAC, audit logs, HITL reviews).
- 3) Model/infra lock-in & cost. Choices made early are hard to reverse. Frugine’s model-agnostic, RAG-first architecture supports deployment in cloud VPC or on-prem, enabling model swaps without app rewrites and aligning costs to real usage.
- On top of these, we provide telemetry dashboards to prove ROI (engagement, retention, conversion) and localization/compliance options for Southeast Asia, China, Japan, and Korea.
- Validation: Frugine is currently piloting with two established enterprises.
Opening the deck…
Page 1
Best customer profile
- Ideal customers: Brands/Retail & CPG, Banking/Insurance, and Media/Entertainment/Telco with digital touchpoints;
- company size 20–20,000+ across SEA, China, Japan, and Korea.
- Buyer/champions: Founders, Product/Growth leads (SMB) and CDO/CTO, Digital/Innovation, Marketing/Brand/CX, L&D and Compliance/IT Security (mid-large).
- Key pain points: (1) Fragmented GenAI workflow; (2) Slow idea→pilot (months); (3) Model/infra lock-in & opaque cost; (4) Need RBAC, auditability, HITL, data residency; (5) ROI visibility on engagement/retention/conversion.
- Why they benefit most: Frugine provides one governed console + APIs/SDKs to standardize a 6–8-week PoC, deploy cloud VPC or on-prem, avoid lock-in, and measure outcomes with telemetry—equally effective for SMBs with limited engineering and for enterprises with strict governance.
Notable paying customers
- Under NDA: One financial services company and one traditional enterprise are currently piloting Frugine; details remain confidential.
Key Milestones & Business Traction Enterprise product launch: Shipped Frugine (frugine.com)—an enterprise AI toolflow & operations console unifying prompt lifecycle/versioning, testing sandboxes, product evaluation, task orchestration, analytics dashboards, and bug tracking, with members/roles (RBAC) and permission controls (audit logs are on the roadmap). B2C platform live: Web platform launched with paid subscriptions and a trial, enabling individuals/small teams to compose custom AI toolflows and smoothly upgrade to the enterprise tier. Customer traction (under NDA): One financial institution and one traditional enterprise are piloting Frugine. Time-to-value: Standardized 6–8-week PoC (idea → pilot), model/infra vendor-agnostic, deployable in cloud VPC or on-prem. Technical breakthroughs: Emotion Engine (real-time affect), persistent memory architecture (profiles + episodic history), hybrid rules engine + generative models, voice/multimodal APIs, and a model-agnostic, RAG-ready stack with telemetry for engagement/retention/conversion. Ecosystem: Accepted into the NVIDIA Inception Program (access to RTX/Hopper GPUs and solution-architect guidance). Funding: Seed round closed (Feb 22, 2025) — SGD 200,000 for 7%; Angel round ongoing — target ~SGD 2.0M (≈US$1.55M) for 7%. Consumer foundation: Built Metlivi, a mobile open-world “healing” social experience powered by the same emotional-AI stack. Regional readiness: Localization and data-residency options for Southeast Asia, China, Japan, and Korea.