Noesis Research

Research that shows its working.

We build the models, benchmarks and positioning frameworks behind early-stage investment and strategy decisions. Every figure arrives with the source it came from and the assumption it rests on.

See what we do About us
50,000+
startup data points modelled
100+
economic indicators in the pipeline
20+
comparable companies scored
15+
precedent transactions benchmarked

What we do

Four pieces of work, each built to be handed over and rerun by someone else.

Signal fusion across startup data

A proprietary scoring and sentiment fusion model applied across 50,000+ startup data points. FRED and yFinance feed it directly, so the signals stay tied to macro conditions instead of floating free of them.

Valuation and forecasting pipeline

A machine learning pipeline built on more than 100 economic indicators, returning startup valuation, revenue forecasts and financial, operational and risk metrics. It runs end to end from a single company input, with no manual step between the query and the output.

Competitive positioning

A 2x2 positioning matrix paired with weighted feature scoring, applied across 20+ comparable companies. It locates the whitespace worth moving into and shows where a defensibility claim does not hold up.

Valuation benchmarking

A benchmarking template built on 15+ precedent transaction comparables, structured to produce multiple ranges that hold up line by line in an early-stage review rather than a single headline number.

How work is delivered

The model, not just the answer

Excel and Python builds are handed over whole, with the workbook, the assumptions and the code that produced the output.

Dashboards when data stays live

Power BI, Tableau or an embedded web view, for work that has to be revisited each month rather than read once.

Written analysis in full

The reasoning laid out step by step, so a reviewer can follow how the conclusion was reached and challenge it where they disagree.

About us

Noesis Research is an independent research practice working with founders, investors and advisory teams.

We take on the analysis that has to survive a second reading. Valuation ranges that get questioned in a committee. Positioning claims that get tested by a competitor. Forecasts that get compared, a year later, against what actually happened.

Most research of this kind arrives as a conclusion with the workings stripped out. We do the opposite. Sources, assumptions and model structure ship with every engagement, so the work can be checked, rerun and extended without us in the room.

Founders

Seyan Sonone
Seyan Sonone
Co-founder
LinkedIn profile
Shashwat Mathur
Shashwat Mathur
Co-founder
LinkedIn profile