Prediction-market infrastructure, built by an operator.

Market-making systems, resolution-risk analysis, and venue integration for trading firms building on Polymarket, Kalshi, and the next wave of event exchanges. Led by Dejan Nešić, sportsbook-platform engineer turned prediction-market operator.

Book an intro call Read the research ↓

Operator, not observer

Designed and operates a live market-making system on Polymarket: TypeScript, maker-first execution, Kelly-capped sizing, combinatorial arbitrage detection

Recalibrated execution economics through the March 2026 fee-regime change

Research on UMA resolution mechanics and cross-venue settlement: Polymarket / Kalshi / ForecastEx

Previously: sportsbook platform engineering at MetricGaming

10+ years senior production engineering: TypeScript, Go, Vue, data pipelines

Services

Three ways to engage

01 Fixed scope · 2 weeks

Venue Entry Audit

from €4,500

For desks evaluating entry. Deliverables:

Category-level fee/rebate economics under current schedules

Oracle and resolution-risk map: UMA and venue-specific rules

Execution and data-architecture review

Written go/no-go memo with quantified assumptions

02 Monthly

Advisory Retainer

from €4,000/mo

Ongoing microstructure counsel:

Resolution-rule reviews and settlement-architecture guidance

Fee/rebate strategy

Hiring and vendor evaluation

Async Slack access plus a weekly call

03 Hourly / fixed-scope

Build Sprints

from €120/h or fixed-scope

Hands-on engineering:

Execution systems and CLOB integrations

Market-data pipelines and risk dashboards

Settlement monitors

TypeScript/Go, production-grade, documented handover

Fixed-scope preferred. Proposals within 48 hours of the intro call.

How engagements run

Intro call

30 minutes, fit and scope. NDA on request.

Proposal in 48h

Fixed scope, fixed price where possible.

Start within two weeks

Weekly written updates. Based in Serbia (CET); daily overlap with US mornings.

brocode.rs/research

Research

July 2026 Resolution risk is the mispriced variable UMA settlement mechanics, with case studies. July 2026 Anatomy of a thin-book loss An honest post-mortem on adverse selection and sizing in prediction-market microstructure. August 2026 Maker/taker math under the March 30, 2026 fee schedule What the new category-tiered fees do to quoting economics.

New research by email. No noise.

Dejan Nešić · Serbia (CET)

About

I'm Dejan Nešić. I spent a decade building production systems, including sportsbook platform engineering at MetricGaming, before designing and operating my own market-making system on Polymarket. The fee tables, oracle edge cases, and thin-book failure modes in my research come from operating capital in these markets, not from reading documentation. BroCode is my Serbia-registered practice; I work remotely with teams worldwide, with daily US-morning overlap.

For

Prop firms and market makers entering event markets

Funds doing venue due diligence

Sportsbook and quant teams crossing over

Prediction-market startups needing senior infrastructure

Not for

Trading signals or tips

Managed accounts

Anything promising returns

FAQ

Do you trade on behalf of clients?

No. Engineering and research consulting only: no managed trading, no performance fees.

Where are you based?

Serbia (CET). Daily overlap with US mornings; calls typically 15:00–18:00 CET.

How do engagements start?

Intro call → scoped proposal within 48 hours → start within two weeks.

Do you sign NDAs?

Yes, routinely.

What's your stack?

TypeScript and Go primarily; Python where it fits; production data pipelines and dashboards.

Availability?

Limited engagements per quarter. Current status:

Available · Q3 2026

If your desk is looking at event markets, the audit pays for itself in avoided mistakes.

Book an intro call or write to dejan@brocode.rs