AI Engineer

Job description

Job Title: AI Engineer

Company
: Darwix AI

Location
: Gurgaon (On-site)

Type
: Full-Time

Experience
: 2-6 Years

Level
: Senior Level

About Darwix AI

Darwix AI
is one of India's fastest-growing GenAI startups, revolutionizing the future of enterprise sales and customer engagement with real-time conversational intelligence. We are building a GenAI-powered agent-assist and pitch intelligence suite that captures, analyzes, and enhances every customer interaction—across voice, video, and chat—in real time.

We serve leading enterprise clients across India, the UAE, and Southeast Asia and are backed by global VCs, top operators from Google, Salesforce, and McKinsey, and CXOs from the industry.

This is your opportunity to join a high-caliber founding tech team solving frontier problems in real-time voice AI, multilingual transcription, retrieval-augmented generation (RAG), and fine-tuned LLMs at scale.

Role Overview

As the
AI Engineer
, you will drive the development, deployment, and optimization of AI systems that power Darwix AI's real-time conversation intelligence platform. This includes voice-to-text transcription, speaker diarization, GenAI summarization, prompt engineering, knowledge retrieval, and real-time nudge delivery.

You will lead a team of AI engineers and work closely with product managers, software architects, and data teams to ensure technical excellence, scalable architecture, and rapid iteration cycles.

This is a high-ownership, hands-on leadership role where you will code, architect, and lead simultaneously.

Key Responsibilities

1.
AI Architecture & Model Development

• Architect end-to-end AI pipelines for transcription, real-time inference, LLM integration, and vector-based retrieval.
• Build, fine-tune, and deploy STT models (Whisper, Wav2Vec2.0) and diarization systems for speaker separation.
• Implement GenAI pipelines using OpenAI, Gemini, LLaMA, Mistral, and other LLM APIs or open-source models.

2.
Real-Time Voice AI System Development

• Design low-latency pipelines for capturing and processing audio in real-time across multi-lingual environments.
• Work on WebSocket-based bi-directional audio streaming, chunked inference, and result caching.
• Develop asynchronous, event-driven architectures for voice processing and decision-making.

3.
RAG & Knowledge Graph Pipelines

• Create retrieval-augmented generation (RAG) systems that pull from structured and unstructured knowledge bases.
• Build vector DB architectures (e.g., FAISS, Pinecone, Weaviate) and connect to LangChain/LlamaIndex workflows.
• Own chunking, indexing, and embedding strategies (OpenAI, Cohere, Hugging Face embeddings).

4.
Fine-Tuning & Prompt Engineering

• Fine-tune LLMs and foundational models using RLHF, SFT, PEFT (e.g., LoRA) as needed.
• Optimize prompts for summarization, categorization, tone analysis, objection handling, etc.
• Perform few-shot and zero-shot evaluations for quality benchmarking.

5.
Pipeline Optimization & MLOps

• Ensure high availability and robustness of AI pipelines using CI/CD tools, Docker, Kubernetes, and GitHub Actions.
• Work with data engineering to streamline data ingestion, labeling, augmentation, and evaluation.
• Build internal tools to benchmark latency, accuracy, and relevance for production-grade AI features.

6.
Team Leadership & Cross-Functional Collaboration

• Lead, mentor, and grow a high-performing AI engineering team.
• Collaborate with backend, frontend, and product teams to build scalable production systems.
• Participate in architectural and design decisions across AI, backend, and data workflows.

Key Technologies & Tools

Languages & Frameworks
:

Python, FastAPI, Flask, LangChain, PyTorch, TensorFlow, HuggingFace Transformers

Voice & Audio
:

Whisper, Wav2Vec2.0, DeepSpeech, , AssemblyAI, Kaldi, Mozilla TTS

Vector DBs & RAG
:

FAISS, Pinecone, Weaviate, ChromaDB, LlamaIndex, LangGraph

LLMs & GenAI APIs
:

OpenAI GPT-4/3.5, Gemini, Claude, Mistral, Meta LLaMA 2/3

DevOps & Deployment
:

Docker, GitHub Actions, CI/CD, Redis, Kafka, Kubernetes, AWS (EC2, Lambda, S3)

Databases
:

MongoDB, Postgres, MySQL, Pinecone, TimescaleDB

Monitoring & Logging
:

Prometheus, Grafana, Sentry, Elastic Stack (ELK)

Requirements & Qualifications

Experience

• 2-6 years of experience in building and deploying AI/ML systems, with at least 2+ years in NLP or voice technologies.
• Proven track record of production deployment of ASR, STT, NLP, or GenAI models.
• Hands-on experience building systems involving vector databases, real-time pipelines, or LLM integrations.

Educational Background

• Bachelor's or Master's in Computer Science, Artificial Intelligence, Machine Learning, or a related field.
• Tier 1 institute preferred (IITs, BITS, IIITs, NITs, or global top 100 universities).

Technical Skills

• Strong coding experience in Python and familiarity with FastAPI/Django.
• Understanding of distributed architectures, memory management, and latency optimization.
• Familiarity with transformer-based model architectures, training techniques, and data pipeline design.

Bonus Experience

• Worked on multilingual speech recognition and translation.
• Experience deploying AI models on edge devices or browsers.
• Built or contributed to open-source ML/NLP projects.
• Published papers or patents in voice, NLP, or deep learning domains.

What Success Looks Like in 6 Months

• Lead the deployment of a real-time STT + diarization system for at least 1 enterprise client.
• Deliver high-accuracy nudge generation pipeline using RAG and summarization models.
• Build an in-house knowledge indexing + vector DB framework integrated into the product.
• Mentor 2–3 AI engineers and own execution across multiple modules.
• Achieve