AI Founding Engineer

AI Founding Engineer ## Job Summary We are seeking a skilled **AI Founding Engineer** to develop and maintain advanced AI systems, primarily focusing on large language models (LLMs) and voice technologies. The role involves building evaluation frameworks, improving generation processes, ensuring data accuracy, and deploying safe, reliable AI solutions that serve real users. The engineer will contribute to creating high-quality, compliant, and dependable AI products, particularly in voice and language understanding domains. ## Responsibilities - Build and maintain **evaluation systems**: create representative test sets, define quality and compliance thresholds, and identify regressions before deployment. - Enhance the **generation stack**: optimize retrieval, prompt design, model selection, verification, routing, and safe human handoff mechanisms. - Ensure **financial facts grounding**: verify that AI explanations are based on approved source data, avoiding the invention of rates, eligibility, premiums, or product terms. - Manage **safe deployment**: oversee versioning, offline evaluation, canary releases, rollback procedures, observability, and maintain an error backlog tied to real user impact. - Improve **voice system dependability**: enhance Automatic Speech Recognition (ASR), endpointing, barge-in, Text-to-Speech (TTS), pronunciation, and reduce time-to-first-audio across Indian languages. - Design **safe workflows**: develop routing, Text-to-SQL, and implement protections against prompt injection, data leakage, and unsupported claims. ## Qualifications - 2+ years of experience building **LLM systems** that serve real users, with a focus on **evaluation harnesses** and regression detection. - Proven experience in creating and explaining **evaluation frameworks** and their impact on system improvements. - Strong **backend ownership**: experience with services, queues, latency optimization, observability, and on-call responsibilities. - Hands-on experience with **real-time voice technologies**: streaming ASR/TTS, Voice Activity Detection (VAD), endpointing, barge-in, and managing latency budgets. - Sound **judgment** in regulated or high-trust environments, with the ability to decide when to block releases. - Educational background in Computer Science, Engineering, or related fields (e.g., BTech, BSc, MTech, or equivalent). ## Preferred Skills - Experience in **fintech, insurance, healthcare**, or **Indian-language NLP**. - Knowledge of **retrieval systems, routing**, and **early-stage AI product development**. - Familiarity with **prompt injection prevention**, **data leakage mitigation**, and **supporting multi-language voice systems**. ## Experience - Minimum of **2+ years** working on **large language models** or **voice AI systems**. - Demonstrated success in deploying **production AI systems**, with experience in handling failures and iterative improvements. ## Environment - The role may involve working in a **hybrid or in-office setting**, depending on company policies. - Focus on **high-trust, regulated environments** requiring careful system design and deployment. - The work involves collaboration across backend services, voice processing modules, and evaluation teams. ## Salary *Not specified.* ## GrowthOpportunities *Not specified.* ## Benefits *Not specified.*