Company Description Varpas is a design and technology company dedicated to creating meaningful value through innovative engineering solutions. We are developing a healthcare-focused SaaS product aimed at digitizing surgery center operations. As an early-stage startup based in Pune (Deccan area), we offer a dynamic work environment that thrives on execution and value creation. If you’re comfortable working in a fast-paced, context-switching environment, we’d love to have you on our team. Role Description We are seeking a talented Full Stack Developer with expertise in web development. You will design, develop, and maintain robust and scalable applications across web and mobile platforms. Collaborating with cross-functional teams, you will define, design, and implement new features while ensuring performance, quality, and responsiveness. Responsibilities Frontend Development: Build responsive and dynamic web pages using Next.js (experience with Angular or Vue.js is a plus). Backend Development: Develop backend services with Node.js and write Python scripts for automation and backend services. API Integration: Integrate RESTful APIs and third-party services. Database Management: Work with both SQL and NoSQL databases, including MongoDB. Version Control: Maintain code quality using Git and version control best practices. Problem Solving: Debug and troubleshoot issues across web and mobile applications. Collaboration: Work with cross-functional teams to deliver high-quality products. Requirements Bachelor’s degree in Computer Science, Engineering, or a related field. Proven experience in full stack web development. Proficiency in HTML5, CSS3, JavaScript, and frameworks like Next.js. Experience with MongoDB; familiarity with SQL and other NoSQL databases. Basic understanding of Python scripting. Basic understanding of system architecture and familiarity with HL7 schema for EMR integrations. Experience with RESTful API integration. Proficient with Git and version control tools. Strong problem-solving skills and attention to detail. Ability to work collaboratively in a cross-functional team environment. Preferred Qualifications Experience with cloud platforms such as Google Cloud (AWS or Azure is a plus). Familiarity with DevOps practices and tools like Docker, Kubernetes, or Jenkins. Understanding of security principles, HIPAA compliance, and best practices in healthcare software development.
Role Description We’re hiring an LLM Engineer to own our conversational AI stack end-to-end—prompt design, RAG, model selection/fine-tuning, evaluation, deployment, and reliability. You’ll build low-latency, high-quality experiences on cloud infrastructure, collaborate closely with product/design, and ship quickly while keeping safety, cost, and maintainability in check. Location: Pune, India Employment: Full-time Responsibilities Architecture & Delivery: Design conversational pipelines (RAG, tools/functions, memory) and take features from prototype to production. Prompt Engineering: Write, version, and A/B test prompts; implement guardrails, system instructions, and tool-use strategies. Model Ops: Evaluate and select models; run SFT/LoRA where appropriate; manage versions, rollouts, and fallbacks. Inference & Performance: Deploy and optimize inference on cloud and specialized accelerators (GPU/TPU/LPU), targeting low latency and predictable cost. Retrieval & Data: Build ingestion pipelines, chunking strategies, embeddings, and metadata for vector search. Quality, Safety & Monitoring: Set up offline/online evals, red-teaming, safety filters, tracing/telemetry, and cost/latency dashboards. MLOps/DevEx: Automate CI/CD for prompts, models, and configs; maintain reproducible environments and strong observability. Collaboration: Work cross-functionally; write clear documentation; mentor teammates on LLM best practices. Requirements 2+ years of software engineering (strong Python ; bonus TypeScript/Node ). 1+ year building production LLM applications (not just POCs). Solid with cloud (preferably GCP: GKE/Cloud Run, IAM, Storage, Pub/Sub, Secret Manager). Practical prompt engineering and RAG experience (LangChain/LlamaIndex or equivalent). Vector databases (Pinecone/Weaviate/FAISS), embeddings, retrieval design. Containers & orchestration (Docker, Kubernetes); APIs (REST/gRPC); infra-as-code (Terraform preferred). Strong software practices: testing, observability, incident response, and performance tuning. Nice to Have Fine-tuning (SFT, LoRA/QLoRA), dataset curation, labeling workflows. vLLM/TGI/TensorRT-LLM; batching, caching, KV-cache optimization; quantization (AWQ/GPTQ). Evaluation tooling (Ragas, DeepEval, Promptfoo) and human-in-the-loop review loops. Realtime/streaming UX (SSE/WebSockets) and speech integrations (ASR/TTS). Security & compliance basics for sensitive data (PII handling, audit logging, RBAC; HIPAA awareness). Healthcare data standards (FHIR/HL7) and EMR/EHR integrations. How We Work Ship small, measure, iterate; ownership over a focused problem area. Tight collaboration with product/design; pragmatic experimentation. Some overlap with U.S. Eastern time for key meetings (a few evenings IST/week). What We Offer Competitive salary with performance bonus and equity. Budget for model/inference/eval tooling and observability. Learning stipend and conference support. High impact—your work will shape a flagship conversational AI product. Application Send your resume, GitHub/portfolio, and a brief note covering: A production LLM system you built (stack, evals, latency/cost outcomes). Your experience deploying and optimizing inference on cloud and accelerators. A tricky prompt/RAG issue you solved and how you measured improvement.