Product

Four layers. One intelligence system.

Starlit is a vertically integrated platform — from raw data ingestion to structured investment output. Each layer compounds the value of the one below it.

01

Data Store

Foundation

02

Data Ingestion

Infrastructure

03

Analysis Layer

Intelligence

04

Dashboard Layer

Interface

01

Data Store

Foundation

What it does

A unified, continuously updated repository that aggregates structured and unstructured data across every source your firm relies on — market data providers, private market platforms, alternative data, news and media, regulatory filings, and your own internal research.

Why it matters

Buy-side teams currently pull data from 8–12 disconnected sources. No single system of truth exists. Analysts spend 60%+ of their time on data gathering rather than analysis. The Data Store eliminates this friction by creating a normalized, queryable foundation.

Buy-side use cases

1

VC analyst pulls real-time revenue proxies alongside internal valuation models for a Series B candidate

2

PE associate benchmarks portfolio company KPIs against sector comps without switching between Bloomberg and PitchBook

3

Hedge fund PM queries alternative data signals alongside traditional fundamental metrics in a single interface

Capabilities

Bloomberg, FactSet, Refinitiv
PitchBook, Preqin, CB Insights
SensorTower, Second Measure, Similarweb
SEC filings, patent databases
News APIs, earnings transcripts
Internal memos, CRM notes, IC decks
02

Data Ingestion

Infrastructure

What it does

A processing layer that normalizes heterogeneous inputs — structured APIs, unstructured documents (PDFs, filings, transcripts), news feeds, and internal notes — into a unified data model. Handles real-time streaming for market data and batch processing for periodic reports.

Why it matters

Raw data is useless without normalization. A revenue figure from PitchBook, a growth estimate from an internal model, and a proxy signal from SensorTower all need to be reconciled, timestamped, and confidence-scored before they can be compared or synthesized.

Buy-side use cases

1

Earnings transcript from a portfolio company is automatically parsed, key metrics extracted, and flagged against your firm's watchlist thresholds

2

Internal deal memo uploaded as PDF is structured into searchable fields and linked to the relevant company profile

3

Alternative data feed from SensorTower is normalized to match your internal demand-proxy framework

Capabilities

API connectors (REST, WebSocket, FIX)
Document parsing (PDF, DOCX, HTML)
NLP extraction from earnings calls
Structured CSV / Excel ingestion
Email and Slack integration
Custom webhook endpoints
03

Analysis Layer

Intelligence

What it does

The core reasoning engine. Encodes your firm's proprietary investment logic (Firm Brain) — sectors, thresholds, red flags, conviction requirements, pattern recognition — and combines it with AI models to generate first-, second-, and third-order insights. Every output is traceable to its source data and reasoning chain.

Why it matters

Investment decisions depend on individual judgment that is rarely codified. When a senior partner retires or a key analyst leaves, institutional knowledge walks out the door. The Analysis Layer makes your firm's best thinking persistent, consistent, and scalable.

Buy-side use cases

1

System applies your firm's 'minimum 90% retention, sub-2x burn' filter across 200 companies and surfaces the 12 that pass — with reasoning for each borderline case

2

AI generates a variant perspective on a consensus-negative name, identifying a catalyst the market is underweighting

3

Second-order analysis flags that a competitor's fundraise could commoditize basic features, increasing the value of your portfolio company's moat

Capabilities

Firm Brain configuration (your logic)
Multi-model AI synthesis (LLM + quantitative)
Proprietary scoring algorithms
Scenario modeling engine
Variant perspective generator
Second-order effect detector
04

Dashboard Layer

Interface

What it does

Structured output surfaces through company-level dashboards, portfolio views, risk monitors, opportunity rankings, and AI-generated memos. Every metric, insight, and recommendation is interactive, drillable, and exportable.

Why it matters

Analysis without presentation is wasted. The Dashboard Layer translates complex, multi-source intelligence into views that match how investment teams actually consume information — dense but scannable, structured but flexible, comprehensive but focused on signal.

Buy-side use cases

1

CIO opens portfolio dashboard before Monday IC meeting and immediately sees which companies triggered risk alerts over the weekend

2

Analyst clicks into a company profile and reviews the AI-generated thesis summary, variant perspective, and key developments — all updated in real time

3

Fund admin exports a portfolio performance report with standardized KPIs for LP quarterly update

Capabilities

Company-level investment profiles
Portfolio health & exposure views
Risk monitor with severity rankings
Opportunity pipeline with scoring
AI investment memo generator
Exportable IC presentation decks

See the full platform in action.

Walk through the architecture with our team and see how each layer applies to your firm.