Software Engineer - AI

We are looking for an AI Specialist who doesn't just "run models" but buildsintelligent systems. You will be at the heart of our innovation product, transforming how enterprises handle voice and document data. This isn't about research papers that sit on a shelfits about implementingAgentic Workflows,RAG architectures, andSmall Language Models (SLMs) that provide real-world value to thousands of users globally.

Key Responsibilities:

• Agentic System Design:Design and implement AI agents that can reason, use tools, and execute multi-step workflows for document intelligence and voice analytics.
• LLM & SLM Orchestration:Beyond just GPT-4, you will work with Open Source models (Llama, Mistral, Phi) and fine-tune them for specific domain tasks using techniques like LoRA/QLoRA.
• Advanced RAG Architectures:Build and optimize Retrieval-Augmented Generation pipelines, focusing on advanced retrieval (chunking strategies, hybrid search, and re-ranking) to eliminate hallucinations.
• Audio & Document Intelligence:Develop and refine pipelines for speech-to-text diarization, sentiment analysis, and multi-modal extraction from complex enterprise documents.
• Production-Grade AI:Quantize and optimize models for cost-effective, low-latency deployment. You ensure the AI "vibe" is fast and responsive.
• Evaluation Frameworks:Implement rigorous evaluation benchmarks (using tools like RAGAS or G-Eval) to measure model accuracy, bias, and performance.

Essential "Vibe" and Skills:

• The "Hacker" Mindset:You are comfortable diving into the latest GitHub repo or research paper on Monday and having a working POC by Wednesday.
• Deep Tech Stack:Strong proficiency in Python and frameworks likeLangChain, LlamaIndex, or Haystack.
• Model Mastery:Hands-on experience with PyTorch/TensorFlow, specifically focused on Transformer-based architectures and fine-tuning.
• Data Engineering for AI:Skilled in data cleansing, vector embeddings, and working with Vector Databases (Chroma, Pinecone, or Weaviate). Vertex AI working knowledge is a plus.
• API & Deployment:Experience deploying models viaFastAPIand managing GPU orchestration (CUDA/Triton).

Minimum Qualifications:

• 3+ years of experience in AI/ML with a portfolio of shipped AI features or products.
• Proven track record of working with LLMs in a production environment (not just notebooks).
• Experience withVision Transformers (ViT)for document layout analysis orSpeech Models(Whisper, etc.) is a massive plus.

Trellissoft, Inc., with over a decade of experience in the IT industry, specializes in delivering innovative and award-winning solutions to help startups and product companies build advanced technologies and modernize their existing platforms. Our flagship offerings include PulseAI360, an AI-driven conversational intelligence platform; DocuVera360, for document automation; and RelyBill, which enhances billing and claims workflows. We collaborate with clients across various industries, including BFSI, telecom, healthcare, and BPOs, providing scalable and impactful AI services. Our commitment to responsible AI design emphasizes fairness, transparency, and compliance, ensuring real-world results and ROI.

We are an equal opportunity employer committed to diversity, equity and inclusion. We welcome applications from candidates of all backgrounds, regardless of gender, age, marital status, disability, caste, religion or sexual orientation. Trellissoft provides reasonable accommodations for applicants with disabilities throughout the recruitment and employment processes. If you require any assistance or adjustment, kindly contact us at hrd@trellissoft.ai