Milky Way AI
Empowering the World's leading FMCG brands and retailers to gain unprecedented marketing insights just by taking photos
Est 2019|Valuation $1.8M|Raised $220K
Empowering the World's leading FMCG brands and retailers to gain unprecedented marketing insights just by taking photos
Published Sep 19, 2026
No explicit, public "limitations" section was located on the official site during this review.
No public product roadmap was found in the materials reviewed.
Independent, large-sample validation is not publicly available. The most detailed operational signal is a partner profile describing pilots and technical stack; prospective buyers should validate few-shot learning and deployment speed claims through representative pilot tests.
Public end-user reviews on mainstream B2B review platforms (G2, TrustRadius, Gartner Peer Insights) are limited for Milky Way AI. Limited review volume constrains independent validation.
The vendor site presents endorsements attributed to company roles, which make positive statements about improved shelf visibility, speed of data collection, and ease of use. These statements are vendor-selected and not independently verified; they lack detailed measurement context (baseline, sample size, measurement method).
Vendor-presented performance claims include reductions in audit time, product recognition accuracy figures, and sales uplift percentages. These should be tested and validated during a pilot under conditions representative of the buyer's stores and SKUs.
A partner-ecosystem profile (MongoDB) describes pilot deployments and technical stack details; this provides operational context but is not a neutral, third-party performance audit.
Glassdoor shows a limited set of employee reviews with mixed observations (positive comments on learning and people, alongside comments noting fast-moving or less stable early-stage dynamics). These are employer reviews and should not be treated as product performance evidence.
The MongoDB profile describes pilot activity and provides operational detail such as a small pilot later scaled to a significantly larger multi-store deployment, and an indicated tech stack (React Native, React, Node.js, Python Flask, TensorFlow, PyTorch, MongoDB Atlas). This material is partner content and offers feasibility signals (pilot progression and technical approach) but does not constitute an independent performance audit or verified ROI study.
A Product Hunt listing outlines intended workflows (offline audits, rapid report generation, KPIs) and device requirements. Product Hunt posts characterize early-adopter positioning rather than enterprise-grade validation.
Vendor marketing materials present sample ROI claims (reductions in audit time, travel cost savings, sales uplift). No detailed ROI methodology, cohort definitions, or independent attestation are provided in public materials; these claims should be validated via pilot projects with documented assumptions.
Indicatively best fit:
Get a comprehensive analysis of Milky Way AI including market position, competitive landscape, adoption trends, and peer benchmarks.
Download full report →Milky Way AI is a B2B computer-vision-driven retail execution and shelf-analytics platform (product name: InstaShelf) that converts store-shelf photographs into near-real-time KPIs and insights for CPG/FMCG brands and retailers—examples include on-shelf availability (OSA), share of shelf, planogram compliance, price/promo compliance, and competitor intelligence. The vendor describes the product as an alternative to manual shelf audits, which the vendor characterizes as slower and more costly.
On the product site, Milky Way AI highlights:
Important naming ambiguity: "Milky Way AI" can be confused with unrelated "Milky Way" products (astronomy content, iOS jailbreak tweak, other software). This report focuses on milkyway.ai / InstaShelf (Singapore-based, participated in a startup accelerator in 2021).
Target users (per vendor / ecosystem sources):
Target buyers:
Positioning: described by the vendor as a computer-vision-based retail platform for shelf intelligence and retail execution with an emphasis on audit speed compared with manual audits.