UI only deployment This deployment hosts only the Athenis frontend. Connect it to your own backend/API instance to enable analysis.

HackOdisha 2025 • Akash GPU Ready

Clarity in every clause

Upload any document and get plain‑language summaries, obligations, and risks in seconds — with clause‑by‑clause breakdowns, realtime streaming, and multilingual support. Built to run on Akash GPUs and scale from hackathon to production.

Reports → Overview

GPU: A100 (Akash)

94%

Legal Clarity Improvements

From Jargon → Clarity

BEFORE

"The Party of the First Part hereby agrees to indemnify, defend, and hold harmless the Party of the Second Part from and against any and all claims..."

"Notwithstanding any provision herein to the contrary, said obligations shall survive termination..."

+ 47 more complex clauses

AFTER

✓ You agree to: Cover any legal costs if someone sues us because of your actions.

✓ Important: This responsibility continues even after the contract ends.

Clear explanations for all clauses

Hackathon Vision

We set out to build a Global Document Simplifier you can clone and deploy on Akash GPU: a Flask API for LLM inference, a lightweight Go ingestion service, and a clean HTML/CSS/JS frontend. It simplifies contracts, flags risky clauses, streams results via SSE, and exports structured reports.

Uploads: PDF/DOCX/TXT → text extraction, chunking, language detection
Tasks: simplify, summarize, risk analysis with heuristic+LLM hybrid
Multilingual: auto-detect + translate when needed
Streaming: token-by-token updates to the browser via SSE
Observability: Prometheus metrics and structured logs

Why Akash?

Akash makes GPU compute accessible and portable. Our containers request a single GPU and run the full stack behind a secure API — cost‑efficient, portable, and ready for production.

• GPU-ready Dockerfile (CUDA 11/12)

• SDL configured to expose Flask on port 8080

• Private networking between Flask and Go microservice

Try Athenis Now

Experience the power of AI-driven legal document analysis

Document Input

Drop your document here

or browse files

Max 20MB

•

PDF, DOCX, TXT

Analysis Options

Analysis Task

Output Language

API Key (Optional) For enhanced features

Simplified Text

Simplified text will appear here after analysis...

Summary

Document summary will appear here...

Risk Analysis

Risk analysis will appear here after document analysis...

Original Text

Original document text will appear here...

We've cracked the code.

Legal documents don't have to be confusing. Our AI transforms complex jargon into clear, understandable language.

Instant Simplification

Transform legal jargon into plain English in seconds, not hours.

Risk Detection

Automatically identify hidden fees, unfair clauses, and potential risks.

Multi-Language

Support for English, Hindi, Spanish, German and more languages.

Export Reports

Download detailed analysis reports in JSON or print-friendly formats.

See the Big Picture

Documents shouldn’t be a barrier. Athenis turns complex contracts, policies, and agreements into clear, actionable insights you can trust.

Every clause is analyzed for obligations, risks, and exceptions — explained in plain language and streamed in real time as the model thinks.

From rental agreements to enterprise contracts, understand exactly what you’re agreeing to, in your language.

Engineered for production from day one: GPU‑backed inference, robust streaming, caching, metrics, and CI/CD — deployable on Akash with minimal setup.

Architecture Overview

Three services, one goal: clarity in every clause.

Flask API (GPU)

Transformers + PyTorch with optional bitsandbytes quantization and CPU fallback.

POST /api/v1/inference
POST /api/v1/stream (SSE)
POST /api/v1/upload
GET /api/v1/health · GET /metrics

Go Ingestion (GoFr)

PDF/DOCX → text, chunking, language detection.

POST /ingest returns chunk metadata
Talks to Flask over internal network
Clean, containerized, tiny footprint

Frontend (HTML/CSS/JS)

No frameworks required; clean, modern UI.

File upload + paste text
Side-by-side original vs simplified
Risk highlighting + drilldown
Export JSON + printable report

Security

API key middleware, hardened headers, token bucket rate limiting.

Observability

Prometheus /metrics, structured logs, example Grafana dashboard.

CI/CD

GitHub Actions: lint, test, build, optional push and deploy.

Akash GPU Deployment

Portable, cost-efficient GPU inference on a decentralized cloud.

Simple Setup Requirements

Choose your GPU power level - from basic to premium options
Set up your service connection and monitoring dashboard
Configure your API keys and model preferences
Enable smart memory optimization for better performance
Secure private networking between your services
Automatic backup system when main service is unavailable

Tip: Use version tags to ensure your deployments work consistently every time.

Quick Launch Guide

Upload your application packages to the cloud registry
Configure your resource requirements and service settings
Set up your security credentials and preferences
Launch your deployment through the simple dashboard
Verify everything is running smoothly and start using your service

• Uses modern deployment technology

• Automatic builds when you update your code

• Built-in performance monitoring included

API Endpoints

Plain‑language insights over HTTP — fast, streaming, and easy to integrate.

POST /api/v1/inference

Input text, choose task (simplify | risk | summarize), select language. Returns overview, plain_language, clause_explanations, risks_detected, meta.

POST /api/v1/stream

SSE endpoint for token streaming and progress events.

POST /api/v1/upload

Multipart upload (PDF/DOCX/TXT) → id + extracted text metadata.

GET /api/v1/health · GET /metrics

GPU memory, model readiness, and Prometheus metrics.

Why Choose Athenis?

See how Athenis replaces hours of reading with instant clarity.

Traditional

• Hours of reading
• Dense legal jargon
• Missed obligations and risks
• Expensive expert reviews
• Language barriers
• No clear highlighting
• Slow, manual analysis

With Athenis

• Instant, streaming analysis
• Plain‑language summaries
• Automatic risk and obligation detection
• Affordable, GPU‑powered inference
• Multilingual support
• Clear, clause‑by‑clause highlights
• Exportable, structured reports

Expert Plus

• AI + human review
• Legal expert validation
• Custom risk profiles
• Priority support
• Advanced analytics and metrics
• Compliance and audit trails
• White‑label options

Map Your Success

Upload

Upload your legal document or paste the text directly into our secure platform.

Analyze and Clarify

Our AI processes the document, simplifies complex language, and identifies potential risks.

Get Clear Results

Receive plain-language summaries, risk assessments, and actionable insights instantly.