PN
Priya NairLondon
StealthEx-Staff Engineer, Stripe · ex-Monzo
- Left Stripe six weeks ago
- New GitHub organisation
- Domain registered
First seen 3 days agoL profile · D dossier
An applied AI lab for venture capital
Your investment process. Your own AI.
Sonexa Labs develops proprietary models and agents around each fund’s investment strategy, data and internal processes. Built for how your team works, with agreed deliverables your firm can retain.
Trusted by venture firms
- KAUSARINVESTMENT GROUP
- ACHARIVENTURES
- KeningfordPARTNERS
- cowboy ventures
- NextView Ventures
01 / The problem
No two funds think alike.
The investment thesis, the questions partners ask, the way analysts work and the firm’s accumulated knowledge are different at every fund. Generic tools start with the same workflow. We start with yours.
01 / Strategy
Different investment theses.
What matters at one fund may be noise to another.
02 / Context
Different institutional knowledge.
Research, decisions and relationships are unique to your firm.
03 / Execution
Different operating processes.
A standard tool rarely reflects every handoff, decision and review.
02 / Inside the lab
We build around the way you invest.
We begin with how your team evaluates opportunities and makes decisions. The systems take shape around that process, not the other way around.
Your approachYour criteriaYour process→FUND SPECIFIC AI
01 / Foundation
Your investment approach sets the direction.
We start by understanding how your team identifies opportunities, builds conviction and makes decisions. That context guides what we build.
02 / Model
Your criteria shape the intelligence.
We translate documented investment requirements into defined model behavior and evaluation standards.
03 / Agent
Your workflow determines the action.
Agents handle scoped tasks, show supporting evidence and pass decisions to investors when judgment is required.
03 / Two ways to work with the lab
Purpose-built systems. Two ways to access them.
01 / Tailored solutions
Engineered around your fund.
Commission models and agents designed for your firm’s internal processes, investment standards and permitted data.
Explore Tailored Solutions ↗02 / Standalone products
Developed in the lab. Ready to use.
Access individual products addressing common venture workflows, starting with Juno for early-stage deal sourcing.
Explore Products ↗04 / Applications
Across the investment lifecycle.
Choose the workflow. We define the required model, agents, data and review points with your team.
01 / Discover
Deal Discovery & Market Intelligence
Build sourcing intelligence around the markets and signals your team cares about.
02 / Evaluate
Screening & Investment Research
Reflect your firm’s screening questions and research standards in models and agents.
03 / Diligence
Due Diligence & IC Preparation
Help teams work through fund-specific diligence and prepare decision materials.
04 / Monitor
Portfolio Intelligence
Make portfolio data and company reporting more useful to the investment team.
05 / Report
Fund Operations & LP Reporting
Design agents around internal reporting and investor communication workflows.
06 / Remember
Institutional Knowledge & Firm Operations
Make past research and investment context usable across the firm.
05 / Ownership & control
A capability your fund can keep.
Not simply another seat in a vendor’s product. Sonexa Labs develops defined systems for your firm, with deliverables, deployment and ownership terms agreed as part of the engagement.
Scope and deliverables defined upfront 01 / Define
Fund data handled within agreed boundaries 02 / Control
Documentation and handover agreed contractually 03 / Retain
06 / First product
Meet Juno. Early signal. Earlier discovery.
Find the company before the fundraise.
Juno brings early company-building activity into an investor’s sourcing workflow. Discover emerging teams, add companies to your radar, enrich the picture and track outreach.
Juno / A product of Sonexa LabsIllustrative UI · sample founders
JunoReview 1–4Reflect RSystem121,304 founders · 10,389 verified · 122,025 companies · 521 signals this week
KEYS1Pass2Punt3Reach out4Remind0LikeJUMPLOpen their profileDFull dossier
Today’s deck 4 / 12Built overnight · re-surfaced first, then punts, then net-new.
AllStealthSignalsPunted
07 / Development approach
From process to working system.
01
Understand
Map the decisions, sources, workflows and desired output.
02
Architect
Agree on model behavior, permissions and human review.
03
Develop + validate
Build and evaluate against real examples and edge cases.
04
Deploy + deliver
Hand over the agreed system, documentation and support plan.
08 / Evidence
Work from the lab. Designed to be measured.
Case 01 / In-house engagement
Kausar Investment Group
Our own investment group, and the lab’s first engagement. Every system we offer runs here on live deal flow before it reaches a client fund.
- Workflow
- Deal discovery → screening → IC preparation
- Baseline
- Sourcing through inbound decks and personal networks. Screening notes spread across inboxes and documents.
- Built
- A sourcing agent tuned to the group’s thesis, a screening model that scores against its written criteria, and IC memo drafts assembled from the research trail.
- Review points
- Partners approve every pass or advance decision. Agents show the sources behind each score.
ResultTwo quarters against the baselineIllustrative
4.5×companies reviewed per week
~40 → ~180
~40 → ~180
45→12minutes per first-pass screen
2 days→same dayto a first IC memo draft
3 of 8IC-advanced deals first surfaced by the sourcing agent
How we document our work
- Engagement
- Workflow
- Baseline
- Result
- Client-approved quote
No invented metrics or customer marks.
A working conversation
What should we build for your fund?
Tell us where your investment process has friction. We’ll examine whether a fund-specific model or agent could address it.