SYRA AI Assistant | Shyam Sunder Kumar

Planned Features

Features Live

45%

Build Progress

AI Model Integrated

SYRA_TERMINAL

syra@python3 ~ assistant_runtime.log

// ─────────────────────────────────────────────────────

// SYRA AI Assistant v0.4.5 · Python 3.x

// Smart Your Response Assistant — Initializing...

// ─────────────────────────────────────────────────────

$ python syra.py --mode=voice --wake="hey syra"

✓ SpeechRecognition engine google.cloud.speech loaded

✓ pyttsx3 TTS engine en-US voice ready

✓ PyAudio mic stream 16kHz · mono open

✓ OpenAI API gpt-4o connected

⚡ Wake-word detector building... in-progress

// ── Awaiting wake-word "Hey SYRA" ──────────────────

🎤 WAKE_WORD_DETECTED → "Hey SYRA" [confidence: 0.99]

// ── Conversation session started ────────────────────

USER › "What's the time right now?"

[STT] Recognized "What's the time right now?" conf:0.97

[NLP] Intent: query_time · Entity: current_time · Local logic

SYRA › "The current time is 3:42 PM."

[TTS] Response spoken via pyttsx3 [en-US]

USER › "Hey SYRA, explain machine learning simply."

[STT] Recognized → routing to OpenAI GPT model...

[API] GPT response received [tokens: 142 · latency: 312ms]

SYRA › "Machine learning is when computers learn patterns from data, not explicit instructions..."

[TTS] Response spoken

// ── Session active — listening ──────────────────────

FEATURES_STATUS

LIVE

🎤

Voice Recognition

Real-time speech-to-text via SpeechRecognition + Google Speech API. Captures mic input, converts to text with 95%+ accuracy.

LIVE

🔊

Text-to-Speech Output

Natural voice output using pyttsx3. Offline-capable, supports speed control, multiple voices — SYRA speaks back naturally.

IN PROGRESS

🔑

Wake-Word Detection

Always-on background listener that triggers only on "Hey SYRA". Custom keyword matching being implemented — like Alexa's hotword engine.

IN PROGRESS

🧠

AI-Powered Intelligent Responses

OpenAI GPT integration for context-aware answers to complex queries. Simple tasks (time, date, calc) handled via local Python logic — complex questions routed to the AI model. Hybrid routing in progress.

PLANNED

⏰

Task Automation

Reminders, calendar, app launching, system control, and custom user-defined voice commands via a plugin architecture.

PLANNED

🌐

Internet Queries

Real-time web search, news headlines, weather, and live factual questions using live internet data retrieval.

VOICE_PIPELINE

🎙️

MIC INPUT

User speaks

🔑

WAKE WORD

"Hey SYRA"

📝

STT

Speech → Text

🧠

NLP / AI

Intent + GPT

⚙️

ACTION

Execute task

🔊

TTS OUTPUT

SYRA speaks

TECH_STACK

// CORE LANGUAGE

Python 3.x

pip Package Manager

Virtual Environment

// SPEECH LIBRARIES

SpeechRecognition

pyttsx3 (TTS)

PyAudio

Google Speech API

Vosk (planned)

// AI & NLP

OpenAI API (GPT)

NLP Processing

NLTK (planned)

spaCy (planned)

// TOOLS

VS Code / PyCharm

GitHub

dotenv (API keys)

DEVELOPMENT_TIMELINE

PHASE 01 — CONCEPT & DESIGN ✓ DONE

Defined SYRA's goals, feature set, and technical architecture. Chose Python as primary language and identified core libraries — SpeechRecognition, pyttsx3, OpenAI API.

PHASE 02 — CORE STT + TTS ENGINE ✓ DONE

Implemented speech-to-text using SpeechRecognition with Google backend and text-to-speech using pyttsx3. Basic voice loop fully working — SYRA can hear commands and speak back responses.

PHASE 03 — WAKE-WORD DETECTION ⚡ IN PROGRESS

Building always-on background thread monitoring audio for "Hey SYRA" trigger phrase. Custom keyword matching being implemented — activates full recognition pipeline only on detection.

PHASE 04 — AI MODEL INTEGRATION ⚡ IN PROGRESS

Integrating OpenAI API for intelligent response generation. Hybrid routing: simple commands handled by local Python logic; complex queries routed to GPT for context-aware answers.

PHASE 05 — TASK AUTOMATION 📋 PLANNED

Adding automation modules — reminders, calendar integration, system control, app launching, and user-defined voice commands via a plugin-style extensible architecture.

PHASE 06 — INTERNET QUERIES & FULL LAUNCH 📋 PLANNED

Web search integration, news, weather, live data retrieval, and final polish — packaging SYRA as a complete installable Python application with GUI configuration dashboard.