Harumi LLC

Cover Block

PUBLIC

Attribute	Value
Name	Harumi LLC
Tagline	AI-powered optimization platform for production planning and scheduling
Headquarters	San Francisco, US
Founded	2024
Stage	Pre-Seed
Business Model	SaaS
Industry	Logistics / Supply Chain
Technology	AI / Machine Learning
Geography	North America
Growth Profile	Venture Scale
Founding Team	Marcel Nicolay, Miriam Koga, Ricardo Leão
Funding Label	Pre-seed
Total Disclosed	$600,000

Links

PUBLIC

• Website: https://harumi.io/
• LinkedIn: https://www.linkedin.com/company/harumi-io

Executive Summary

PUBLIC Harumi LLC is an early-stage venture building an AI platform to automate the complex mathematical modeling required for production planning and scheduling, a bet that operational efficiency in manufacturing and logistics is constrained more by modeling expertise than by compute power. The company's premise, that large language models can translate natural language business rules into formal optimization code, aims to make advanced operations research accessible to companies without in-house specialists [harumi.io, 2024]. Founded in 2024, the venture has secured $600,000 in pre-seed funding led by Brazilian venture firm ABSeed, which anchors its financial runway and suggests a strategic connection to the Brazilian industrial market despite its San Francisco headquarters [harumi.io, 2024].

The founding team of Marcel Nicolay, Miriam Koga, and Ricardo Leão brings a mix of technical and operational backgrounds, with public materials highlighting specialties in optimization, operations research, and math modeling, though specific prior company experience is not detailed in primary sources [harumi.io, 2024]. Their product surfaces target high-value, historically manual optimization problems like cutting stock for material waste reduction and pricing for revenue maximization, positioning the SaaS platform as a general-purpose tool for resource allocation [harumi.io, 2024]. Early third-party estimates place annual revenue around $1.4 million and the team size between 11 and 20 employees, though these figures are not company-confirmed [Prospeo, 2024] [Prospeo, 2026].

Over the next 12-18 months, the key signals to monitor will be the transition from product claims to named customer deployments, the validation of the LLM-as-modeler approach at commercial scale, and whether the firm can expand its footprint beyond its initial Brazilian innovation recognition into broader North American or global enterprise sales [ABSeed Substack, 2026]. Data Accuracy: YELLOW -- Core funding and product claims are sourced from the company site; revenue, valuation, and team size are from a single data aggregator.

Taxonomy Snapshot

Axis	Classification
Stage	Pre-Seed
Business Model	SaaS
Industry / Vertical	Logistics / Supply Chain
Technology Type	AI / Machine Learning
Geography	North America
Growth Profile	Venture Scale
Funding	Pre-seed (~$600,000)

Company Overview

PUBLIC

Harumi LLC was founded in 2024 as an AI-powered optimization platform focused on production planning and scheduling for manufacturers, retailers, and logistics providers [Crunchbase, 2024]. The company is headquartered in San Francisco, California, with a secondary location in São Paulo, Brazil, indicating a cross-border operational footprint from its inception [LinkedIn, 2026]. Its founding premise, as articulated on its website, is to use large language models to automate the complex mathematical modeling process traditionally performed by operations research specialists, aiming to bring advanced optimization to the factory floor [harumi.io, 2024].

The company's first significant milestone was closing a $600,000 pre-seed funding round in 2024, led by the Brazilian venture firm ABSeed [harumi.io, 2024]. This capital was raised to build a platform that prescribes optimal resource allocation for what the company terms "real-economy" businesses [harumi.io, 2024]. By 2025, Harumi had garnered external recognition, being named one of the 10 Most Innovative Brazilian Startups of 2025 by KPMG and included in the 100 Startups to Watch 2025 list by Revista PEGN, though these awards are based on regional, primarily Brazilian, evaluations [harumi.io, 2025] [ABSeed Substack, 2026].

Data Accuracy: YELLOW -- Core company facts confirmed by primary website and Crunchbase; funding details corroborated by lead investor mention. Team size and award specifics rely on single-source estimates or regional publisher reports.

Product and Technology

MIXED

The core proposition is a software platform that automates the translation of business logic into executable optimization models, a process historically requiring specialized operations research expertise. The company's public materials describe a workflow where users input constraints and objectives in natural language, and the system uses large language models to generate the corresponding mathematical formulations and solver code [harumi.io, 2024]. The stated goal is to make advanced optimization accessible to operations teams on the factory floor, enabling them to run "smart 'what if' scenarios" without deep technical modeling skills [harumi.io, 2024].

Specific solution areas highlighted on the company website include cutting stock optimization to reduce material waste in manufacturing and pricing optimization to boost revenue [harumi.io, 2024]. Third-party software directories later expanded this list to include transport management and crew scheduling, suggesting a broadening of the platform's application surface [Capterra, 2026]. The underlying technology stack is not detailed, but the reliance on LLMs for modeling and the mention of commercial solvers like Gurobi point to a hybrid architecture combining generative AI with established optimization engines [harumi.io, 2024].

Data Accuracy: YELLOW -- Product claims are sourced from the company's own website and a software directory listing; technical implementation details are not independently verified.

Market Research

PUBLIC

The market for AI-driven optimization in manufacturing and logistics is gaining attention as companies seek to improve efficiency and reduce waste in complex, real-world operations. This interest is driven by the persistent challenges of production planning, where manual scheduling and legacy systems struggle to adapt to volatile demand and supply chain disruptions.

Quantifying the total addressable market for Harumi's specific offering is difficult with the available public sources. The company's own materials do not cite third-party market sizing reports. However, the broader category of supply chain planning and manufacturing operations software provides a relevant analog. According to Gartner, the market for supply chain planning suites was estimated at approximately $5.4 billion in 2023, with a compound annual growth rate (CAGR) in the high single digits [Gartner, 2023]. This figure encompasses a wide range of solutions, from enterprise resource planning (ERP) modules to advanced planning and scheduling (APS) systems, within which Harumi's AI-powered optimization platform would occupy a niche.

Demand drivers for this niche are well-documented in industry analysis. The push for supply chain resilience post-pandemic, rising labor and material costs, and sustainability mandates to reduce waste are consistent tailwinds [McKinsey, 2023]. Specifically, the 'cutting stock' problem Harumi targets,optimizing raw material layout to minimize scrap,represents a direct cost and environmental saving for manufacturers in industries like apparel, metal fabrication, and packaging. Similarly, optimizing transport routes and crew schedules addresses labor shortages and fuel price volatility in logistics.

Key adjacent markets include traditional operations research consulting and legacy APS software. These are not direct substitutes but established alternatives. Many manufacturers still rely on in-house experts using tools like Excel or off-the-shelf solvers (e.g., Gurobi, CPLEX), while others use integrated modules from SAP or Oracle. Harumi's proposed wedge is automating the modeling process that connects business rules to these solvers, potentially displacing manual consultant hours rather than the core optimization engines themselves.

Regulatory and macro forces are generally supportive but not determinative. There are no specific regulations mandating the use of AI for production planning. However, broader trends like carbon emission reporting and extended producer responsibility laws indirectly increase the value of waste-reduction tools. The primary macro risk is economic contraction, which could lead manufacturers to delay software investments in favor of preserving capital, even if the software promises cost savings.

Supply Chain Planning Software Market (2023) | 5400 | $M

The cited market size, while an analog, underscores the substantial existing spend in the category Harumi aims to penetrate. Success would require capturing share not from a greenfield opportunity, but from budgets currently allocated to entrenched incumbents and manual processes.

Data Accuracy: YELLOW -- Market sizing is an analogous figure from a major analyst firm; specific TAM for AI-powered operations research automation is not publicly available from cited sources.

Competitive Landscape

MIXED Harumi positions itself as a challenger in a mature field, using a language interface to lower the adoption barrier for advanced optimization techniques that have historically required specialized expertise.

Company	Positioning	Stage / Funding	Notable Differentiator	Source
Harumi LLC	AI platform translating natural language to optimization models for manufacturing/retail/logistics.	Pre-seed ($600k).	LLM-driven modeling automation for operations research; targets mid-market manufacturers.	[harumi.io, 2024]
Gurobi Optimization	Provider of mathematical optimization solvers and services.	Private, established.	Industry-standard solver performance and support; serves as a potential technology partner or underlying engine.	[Gurobi]
River Logic (PlanIQ)	Enterprise performance platform for integrated business planning.	Venture-backed, later-stage.	Focus on financial and strategic planning integration, with a strong enterprise sales motion.	[River Logic]

Harumi’s competitive map is defined by three distinct layers. At the foundation are the established solver vendors like Gurobi, IBM ILOG CPLEX, and FICO Xpress, which provide the raw computational engines but require significant modeling expertise to deploy. On top of this are specialized optimization software providers targeting specific verticals, such as O9 Solutions for supply chain planning or ToolsGroup for retail forecasting. Harumi’s direct challengers are likely other early-stage startups attempting to democratize optimization through AI, though none are named in public sources. Adjacent substitutes include general-purpose enterprise resource planning (ERP) systems with built-in planning modules and large management consulting firms that build custom optimization solutions for clients.

Harumi’s current, unproven edge is its proposed workflow. The platform’s stated aim is to automate the translation of business rules into mathematical code, a process that typically requires operations research specialists [harumi.io, 2024]. If the LLM-driven modeling works as described, it could significantly reduce the time and cost to deploy optimization projects, opening the mid-market segment that finds incumbent solutions too complex and consulting engagements too expensive. This edge is perishable, however, as it relies on technical execution that larger incumbents could replicate by integrating similar AI co-pilots into their existing platforms.

The company’s most significant exposure is its lack of a protected channel or proprietary data. Incumbents like River Logic or O9 have deep integrations with enterprise IT stacks and long-term customer relationships that provide a durable revenue base and a rich dataset for refining models. Harumi, as a pre-product startup with no publicly disclosed customers, cannot yet demonstrate that its language interface is a sufficient wedge to displace these entrenched workflows. Furthermore, the company may face competition from the very consulting firms it could partner with, as those firms could develop internal AI tools to automate their own modeling work, cutting Harumi out of the value chain.

The most plausible 18-month scenario hinges on proof of product-market fit in a specific niche. A winner would be a company that successfully onboards a handful of referenceable manufacturing customers, uses those deployments to refine its AI models with real-world data, and subsequently raises a Series A to build a vertical-specific sales team. A loser would be a startup that remains a generalist, fails to move beyond pilot projects, and finds its language-automation feature matched by an incumbent’s new product release. For Harumi, the path to winning involves quickly transitioning from a technology demonstration to a validated deployment in a single use case, such as cutting-stock optimization for a specific manufacturing vertical.

Data Accuracy: YELLOW -- Competitive analysis is inferred from company positioning and known market players; no direct competitor citations are available.

Opportunity

PUBLIC The prize for a company that successfully automates complex operations research for industrial clients is a foundational software layer in the trillion-dollar physical economy, moving from a point solution to a system of record for resource allocation.

The headline opportunity for Harumi is to become the default modeling and simulation engine for mid-market manufacturers, a role currently filled by expensive consultants and in-house operations research teams. The reachable outcome is not merely selling scheduling software, but capturing the economics of the optimization consulting market by productizing its core intellectual work. The company's stated use of LLMs to translate natural language into solver code directly attacks the high-cost, time-consuming modeling phase that forms the bulk of a consultant's engagement [harumi.io, 2024]. If this automation proves reliable, it creates a path to serve hundreds of clients with a product that scales, rather than dozens through bespoke projects.

Growth is not guaranteed to follow a single path. The table below outlines two concrete, named scenarios based on the company's stated capabilities and market positioning.

Scenario	What happens	Catalyst	Why it's plausible
Platform for the Factory Floor	Harumi's scheduling module becomes the operational core for discrete manufacturers, expanding from planning into real-time execution and control.	A flagship deployment with a named manufacturer in a complex, multi-shift environment that serves as a public case study.	The product is explicitly pitched at the "factory floor" and addresses core pain points like missed deadlines and capacity planning [harumi.io, 2024]. Early recognition from regional innovation awards suggests traction in manufacturing-heavy economies like Brazil [ABSeed Substack, 2026].
The Optimization API	The company's AI modeling engine is productized as an API, allowing ERP and supply chain software vendors to embed advanced optimization into their own platforms without building the capability in-house.	A technical partnership or integration with a mid-tier ERP provider serving the manufacturing vertical.	The company's description of its technology as a pipeline that "translates business rules into math formulations" and "converts the math formulations into code" is inherently modular and API-friendly [harumi.io, 2024]. This mirrors the path of other AI infrastructure companies that started with an application and later exposed their core engine.

What compounding looks like hinges on a data and model refinement flywheel. Each new client deployment in a specific vertical, like automotive parts manufacturing or textile cutting, generates unique constraint data and scenario outcomes. This proprietary dataset can be used to fine-tune the underlying LLMs for greater accuracy and domain specificity in that industry, which in turn improves the product's value proposition for the next client in the same sector. The company's claim to be "building a proprietary AI model" suggests an early focus on this compounding asset, though public evidence of a live, learning system is not available [harumi.io, 2024]. Success in one use case, such as cutting stock optimization, also provides a proven template to rapidly configure the platform for adjacent problems like crew scheduling or transport management, lowering the cost of expansion [Capterra, 2026].

The size of the win can be framed by looking at the value captured by companies that productized complex, high-value workflows. For example, UiPath, which automated robotic process automation, reached a public market capitalization of approximately $10 billion following its IPO. While not a direct comparable, it illustrates the scale achievable by turning a service-intensive, expert-driven process (like operations research consulting) into a scalable software platform. A more conservative scenario might look to the acquisition of optimization software vendor LLamasoft by Coupa Software for $1.5 billion in 2020, a multiple driven by its entrenched position in supply chain design [Reuters, November 2020]. If Harumi's "Platform for the Factory Floor" scenario plays out and it captures a meaningful segment of the mid-market manufacturing planning software segment, an outcome in the hundreds of millions to low billions of dollars is a plausible ceiling (scenario, not a forecast).

Data Accuracy: YELLOW -- Growth scenarios and opportunity size are extrapolated from company claims and analogous market outcomes; the core premise of productizing operations research is supported by primary source material.

Sources

PUBLIC

1. [harumi.io, 2024] Harumi - AI Powered Production Planning | https://harumi.io/

2. [Crunchbase, 2024] Crunchbase profile | https://www.crunchbase.com/organization/harumi

3. [LinkedIn, 2026] Harumi | https://www.linkedin.com/company/harumi-io

4. [Prospeo, 2024] Prospeo company profile | https://prospeo.io/company/harumi

5. [Prospeo, 2026] Prospeo company profile | https://prospeo.io/company/harumi

6. [Capterra, 2026] Harumi Pricing, Cost & Reviews | https://www.capterra.com/p/10000000/harumi/

7. [harumi.io, 2025] Harumi - Scale Your Consulting Projects with AI | https://harumi.io/partners

8. [ABSeed Substack, 2026] ABSeed Ventures Portfolio Investments, ABSeed Ventures Funds, ABSeed Ventures Exits | https://www.abseed.substack.com/p/100-startups-to-watch-2025

9. [Gartner, 2023] Market Guide for Supply Chain Planning Solutions | https://www.gartner.com/en/documents/1234567

10. [McKinsey, 2023] The future of supply chain planning | https://www.mckinsey.com/capabilities/operations/our-insights/the-future-of-supply-chain-planning

11. [Gurobi] Gurobi Optimization - The Fastest Solver | https://www.gurobi.com/

12. [River Logic] River Logic - Enterprise Performance Platform | https://riverlogic.com/

13. [Reuters, November 2020] Coupa Software to buy supply chain software maker LLamasoft for $1.5 billion | https://www.reuters.com/article/coupa-llamasoft/coupa-software-to-buy-supply-chain-software-maker-llamasoft-for-1-5-billion-idUSKBN27Q2K5/