Investor Deck · Midas Analytics Limited

See what's next in Asian markets, before everyone else.

The #1 Asia market-intelligence platform — AI that reads Asia's filings, media and signals in their own languages, and answers with sources. Revenue since year one.

founded 2021 · HKUST spin-off US$1.1M+ cumulative revenue HK ICT Award 2025 customers incl. UBS · FedEx · Logitech

Private & confidential

Michele De Filippo, PhD

Founder · Midas Analytics

The problem

Asia is where the growth is. Reading it is broken.

The signal is fragmented and non-English

Filings on CNINFO, EDINET and local registries; narratives breaking in Mandarin, Japanese and local media — hours or days before English summaries reach global desks. By then the entry point is gone.

Incumbent tools weren't built for this

Legacy terminals cost tens of thousands per seat and stay English-first on Asia's long tail; generic AI answers from uncontrolled web sources no professional can put in front of a committee.

Human attention doesn't scale

Analysts burn 5–10 hours a week on initial idea scoping alone — chasing sources, translating, re-checking. Asia produces more disclosure than any team can read.

The product · live capture

One platform: search, ask, and always-on agents.

app.midasanalytics.ai

Automation + AI + human curation = what matters · source-locked, fully traceable

Market Research — Asia-wide search, native languages AskAI — cited, research-grade answers AI Agents — 24/7 monitoring, alerts & briefs

How customers use it · across the trade lifecycle

From idea to exit — and beyond finance.

Pre-trade

Idea generation & market scanning

24/7 thematic agents over multi-language sources — 5–10 hours saved per analyst per week on idea scoping.

Event, policy & supply-chain anticipation

Guidance changes, export controls, port closures — flagged before English summaries move spreads.

Single-name deep diligence

Issuer dossiers across filings, media and regulators — time-to-diligence from days to hours.

Post-trade

Position & risk monitoring

Portfolio-linked agents with severity-scored alerts — "what changed since we put this trade on?", cited.

Performance review & client reporting

Traceable event timelines that explain why a trade worked — exportable to IC memos and letters.

Investor, LP & stakeholder comms

Cited narratives for PM letters and LP updates — consistent storytelling, less senior time.

Same engine serves corporate strategy & supply-chain teams — that's how FedEx and Logitech ended up next to UBS on the customer list.

Why now

Three curves crossing at once.

Geopolitics demands Asia visibility

As supply chains and capital flows rewire, corporate and institutional interest in Asian economies is rising — while local-language opacity gets more costly to ignore.

GenAI made reading-at-scale possible

Machines can finally read everything — but professionals need trusted, source-locked answers, not plausible text. Curation is the differentiator, and we've built it for five years.

Buyers want out of terminal pricing

A US$44B market-data bill is under open pressure from the institutions paying it — desks are actively hunting for smarter, cheaper intelligence layers.

Market size · sourced

A US$44B spend pool, tilting our way.

TAMUS$44.3Bglobal financial market data & analysis spend, 2024

SAM≈ US$9BAsia-Pacific slice (~20%)

WedgeAI-native
Asia intel

TAM — US$44.3B and still growing

Global spend on financial market data & analysis rose 6.4% to a record in 2024. Burton-Taylor via Finextra

SAM — the Asia-Pacific layer

≈20% of global market-data spend sits in APAC (internal estimate on Burton-Taylor base) — the least served by English-first tools, and our home turf.

And the wedge is the fastest-growing spend in finance

Alternative & AI-driven data: US$7.2B (2023) → US$135.7B by 2030 (52% CAGR); Asia-Pacific growing fastest at ~56%. Grand View Research

Market-intelligence software itself: US$12.4B (2025) → ~US$29B by 2033 at 11.2% CAGR (Market Mind Partners). Sizing sources linked; SAM split is an internal estimate.

Traction · logos that don't need captions

Blue-chip customers, on both sides of the platform.

Financial services

UBSglobal bank Scopiainvestment management Crescent Parkinvestments DMOcapital Mordrilcapital

Corporates & institutions

FedExlogistics Logitechconsumer tech HKUST Business Schoolacademia Tai Kwunculture & heritage …and more

7,400+

Asian companies tracked, resolved & deduplicated

~9,000

articles parsed daily, hourly cycles, native languages

24/7

AI agents monitoring watchlists & portfolios

5 yrs

of revenue — customers have paid since year one

Financials · actuals, US$ · 2026 = year-to-date

Bootstrapped to US$1.1M+ cumulative revenue.

RevenueExpenses

~80%

revenue CAGR, 2021 → 2025

break-even+

or better, every single year — growth funded by customers, not burn

33%

operating margin on 2026 revenue to date

Business model

Recurring platform revenue, services as the wedge.

01 · Platform subscriptions

Seats & teams, annual

The core: recurring access to Market Research, AskAI and AI Agents — priced far below a terminal seat, sold to desks and strategy teams.

02 · Enterprise & API

Data where clients live

Enterprise deployments and the Auris API — Midas signals streamed into client stacks, models and workflows. Higher ACV, deeper lock-in.

03 · AI solutions

The paid discovery channel

Bespoke AI engagements for enterprises — the revenue that bootstrapped the company, and a proven pipeline into platform accounts.

Land with a team subscription or an engagement → expand to enterprise → lock in with agents, watchlists and API integrations that compound in value the longer they run.

Competitive landscape

The top-right corner is empty — except us.

→ Asia-native coverage depth → AI-native workflow

Legacy terminals

Deep global data, US$25k+/seat, workflow from the 1990s. English-first on Asia's long tail.

Global AI research platforms

AI-native, but built on English filings & transcripts — Asia local sources are the blind spot.

Local data vendors

Country-deep feeds and terminals, but single-market and with no AI reasoning layer.

Midas

Asia-native sources in native languages and an AI-native, source-locked workflow — at a price a whole team can adopt.

Why it's defensible

The moat compounds daily.

A five-year proprietary pipeline

Own scrapers across 70 sources, a Bloomberg-verified entity backbone, a 4-level topic taxonomy, six-model AI engine — engineering no prompt can replicate.

Data that compounds

Every day the pipeline reads ~9K Asian articles and distills them into hundreds of curated, entity-resolved, topic-tagged additions to a corpus competitors would need years to rebuild.

Switching costs that grow with use

Watchlists, tuned agents, saved searches and API integrations accumulate — the longer a desk runs Midas, the more it would cost to leave.

Institutional trust, certified

ISO 27001 certified; a HKSAR-designated research institution (DLRI); source-locked answers by architecture — the compliance story enterprise buyers require.

Validation · awards, ecosystem, media

Six awards in three years, and counting.

Awards

HK ICT Award 2025 — Best Software / App (ICT Startup Award) Huawei — FinTech category winner · Top 15 Startups in Asia 2023 VivaTech 2025 — Top 12 Fintech Startups (Paris) Finance Innovation — Best Asian Startup TSSSU-O — winner 2023/24 and 2024/25 Agorize — Fintech Fuel winner

Ecosystem

Incubated at HKSTP — Hong Kong's #1 tech startup incubator Portfolio company of the Hong Kong AI Laboratory Spin-off from the HKUST Entrepreneurship Centre Accelerated at Creative Valley (France) — the European bridgehead DLRI-designated by the HKSAR Innovation & Technology Commission

In the press

Hong Kong Economic Times — VivaTech coverage of the platform & AI agents Wen Wei Po — Huawei Cloud APAC startup finals Sing Tao Headline — European market reception …and more

Recognition converts: awards and expos have consistently fed the sales pipeline — a working acquisition channel, not wall decoration.

Go-to-market

Asia-out: Hong Kong home base, Europe beachhead.

Direct enterprise · HK & Asia

Founder-led sales into funds, banks and corporates from Hong Kong — where the customers and the data both live. UBS-to-FedEx logo base as reference sellers.

Europe via distribution

A French distribution partnership signed ahead of VivaTech 2025 and showcased there, plus the Creative Valley network — selling Asia visibility to the desks furthest from it.

Ecosystem & product-led

Awards, media and the HKSTP/HKAI Lab network feed inbound; live-demo trials on the prospect's own coverage convert it.

Team

Built by the people who build it.

Michele De Filippo, PhD

Founder

PhD; founded Midas out of the HKUST ecosystem in 2021 and has run it revenue-funded ever since — product, AI and enterprise sales, from UBS to FedEx.

Three teams, clean seams

Data team — scrapers, enrichment fleet, entity backbone, data-quality SLAs AI team — the six-model engine, AskAI/RAG serving, AI agents Software team — the platform, backend and enterprise integrations

Engineering-led and Hong Kong-based, with university-grade R&D roots (TSSSU ×2, DLRI designation).

Full org chart, cap table and hiring plan live in the data room.

The ask

Five years of proof. Now we pour fuel on it.

We've built the product, the pipeline and the customer base on our own revenue. We're now opening our first institutional round to move from founder-led sales to a repeatable growth engine.

Use of funds · 01

Go-to-market

Sales & customer success in HK/Singapore; activate the European distribution channel properly.

Use of funds · 02

Product & AI

Scale AI agents, ship the Auris API to GA, build out the risk-signal layer already designed into the engine.

Use of funds · 03

Data moat

Widen source coverage across more Asian markets and languages — the corpus is the compounding asset.

Round size and terms are in discussion with early participants — write to michele@midasanalytics.ai for the data room.

Thank you.

See what's next in Asian markets — before everyone else.

midasanalytics.ai

hello@midasanalytics.ai +852 6734 8993 Innocentre · Kowloon Tong · Hong Kong