Inspiration
I was inspired by the complexity of professional League of Legends preparation. Behind every competitive series are hours of match analysis, opponent scouting, champion-pool research and draft preparation. I wanted to create a platform that brings these fragmented workflows together and helps coaches turn historical match data into better strategic decisions. My goal was to build more than another draft recommendation tool: I wanted to create an intelligence system that understands how one game can fundamentally change the possibilities of the next.
What it does
SERIESSIGHT is an esports coaching intelligence platform designed for League of Legends teams, coaches and analysts.
It transforms match data into actionable opponent intelligence through:
- Opponent Scouting: Analyzes champion preferences, pick-and-ban tendencies, team patterns and side-specific strategies.
- Macro Intelligence: Organizes available match statistics, objective control and gameplay patterns into understandable insights.
- Draft Command: Provides an interactive environment for managing picks, bans, champion availability and team compositions.
- Fearless Series Lab: Simulates multi-game series and automatically applies champion restrictions across subsequent games under configurable Fearless Draft rules.
- Evidence Ledger: Connects recommendations to their underlying statistics, sample sizes, data sources and confidence levels.
The goal is to help coaching staff understand not just what an opponent does, but why it matters and how to prepare for it.
How I built it
I approached SERIESSIGHT as a complete coaching intelligence platform rather than a collection of disconnected analytics tools.
I designed the architecture around a modular application with a dedicated data integration layer, normalized match data, analytical services and a deterministic draft engine.
The technical direction uses Next.js, React, TypeScript and Tailwind CSS for the application, with PostgreSQL for structured persistence and validation systems for reliable data handling.
I designed the data layer around GRID's esports data APIs and Riot's Data Dragon static champion information, with a separate fixture-data mode so the platform can demonstrate its workflows without requiring live API credentials.
For the intelligence engine, I focused on reproducible statistical calculations, champion tendencies, sample sizes, composition analysis and evidence-backed recommendations.
I also developed a custom visual direction, including an original identity, dark analytical interface, custom glyphs, interactive visualizations, motion effects, loading animations and a five-game series visualization.
JetBrains WebStorm and Junie form part of the intended development workflow, supporting implementation, refactoring, testing and engineering documentation.
Challenges I ran into
One of the biggest challenges was making Fearless Draft logic reliable across an entire series. A champion picked in an earlier game can affect availability in later games, and different tournament rulesets can apply different restrictions.
I needed to approach this as a deterministic state-management problem rather than simply removing champions from a visual selection screen.
Another challenge was data quality. Match histories can contain missing fields, different patch versions and incomplete statistical coverage. I wanted the platform to distinguish verified observations from incomplete information instead of presenting misleading conclusions.
I also had to balance an ambitious visual experience with usability and performance. Advanced visualizations, animations and 3D effects are valuable only when they help communicate information without distracting from the actual coaching workflow.
Accomplishments that I'm proud of
I'm particularly proud of the direction of the Fearless Series Lab. It makes the consequences of draft decisions visible across an entire series rather than treating every game as an isolated event.
I'm also proud of the evidence-first philosophy behind SERIESSIGHT. Recommendations should not simply appear because an algorithm produced a score. They should have an explanation, supporting observations, sample sizes and identifiable limitations.
Other highlights include:
- Bringing opponent scouting, macro analysis and draft preparation into one product concept.
- Designing configurable Hard Fearless and Soft Fearless rules.
- Creating a dedicated evidence and data-provenance architecture.
- Developing an original esports intelligence visual identity.
- Designing the application around real coaching workflows rather than generic AI interactions.
What I learned
I learned that useful AI-powered analytics requires much more than connecting a language model to a dataset.
Reliable data normalization, statistical reasoning, deterministic business logic and explainability are essential. A recommendation is only useful when the person receiving it can understand where it came from and how much confidence to place in it.
I also learned how important state management becomes when building systems where one decision changes the possibilities of subsequent decisions.
Most importantly, I learned that good product design should make complex information easier to understand and act upon, not simply make it look impressive.
What's next for SERIESSIGHT
I want to take SERIESSIGHT beyond the hackathon and develop it into a mature esports intelligence product.
My next priorities are expanding authorized tournament data integrations, strengthening historical opponent profiling, improving macro analysis, developing more advanced counterfactual simulations and adding comprehensive self-scouting capabilities.
I also want to introduce collaborative coaching workspaces, saved preparation strategies, historical backtesting and more sophisticated recommendation models that can be evaluated against actual outcomes.
Long term, my vision is to make SERIESSIGHT a unified intelligence and preparation platform that helps competitive esports organizations understand their opponents, evaluate their strategies and prepare for every game of a series with greater precision.
SERIESSIGHT. The series, in focus.
Built With
- drizzle-orm
- framer-motion
- github
- grid-api
- jetbrains-webstorm
- junie
- next.js
- playwright
- postgresql
- react
- react-three-fiber
- recharts
- riot-data-dragon
- tailwind-css
- tanstack-query
- three.js
- typescript
- vitest
- xstate
- zod
- zustand