Applied AI Engineer — Memory & Context Systems (Aiva)
Aivar Innovations Private Limited7 hours ago
Coimbatore, IndiaMid Level
Responsibilities
- Build LLM-powered memory extraction pipelines for conversations, documents, and process events.
- Design memory consolidation, progressive summarization, salience scoring, conflict resolution, and forgetting policies.
- Develop context assembly logic that selects, ranks, and compresses memories within task and token constraints.
- Own evaluation infrastructure including retrieval benchmarks, golden datasets, LLM-as-judge evaluations, contamination and hallucination tests, regression suites, and A/B measurements.
- Prototype and productionize embedding, chunking, hybrid sparse-and-dense retrieval, and re-ranking strategies.
- Partner on fine-tuning and efficient deployment of small models for high-volume extraction, classification, and summarization tasks.
- Implement PII detection and redaction within AI governance policies.
- Track memory architecture research and translate relevant approaches into production capabilities.
Requirements
- At least 4 years of software engineering experience with strong Python skills.
- At least 2 years building production LLM or NLP applications.
- Hands-on experience with prompt engineering, structured output extraction, RAG pipelines, and embedding models in real-user systems.
- Experience building evaluation frameworks for LLM systems, including golden datasets, LLM-as-judge patterns, regression suites, and offline/online metrics.
- Strong fundamentals in retrieval metrics, ranking, classification, and statistics for A/B testing.
- Working knowledge of AWS ML services such as Bedrock and SageMaker, or equivalent cloud ML platforms.
- Strong software craftsmanship, including testing, versioning prompts/models/datasets, and reproducible pipelines.
- Experience fine-tuning small or open models with LoRA, QLoRA, or distillation is preferred.
- Familiarity with agent memory frameworks, GraphRAG, MemGPT-style architectures, and reflection or consolidation patterns is preferred.
- Knowledge-graph construction experience with NER, relation extraction, or entity linking is preferred.
- Multilingual NLP experience with Hindi, Tamil, or other Indic languages is preferred.
- Published work, open-source contributions, or strong writing on applied LLM systems is preferred.
Benefits
- Zero-to-one ownership of a platform layer positioned as a differentiator for Aiva’s accelerator suite.
- Work shipped to large regulated enterprise customers across telecom, BFSI, healthcare, and manufacturing at production scale.
- Small, senior team with direct access to founders and customers.
- Competitive compensation with meaningful ESOPs.
- Founded by former AWS/Amazon executives and backed by Bessemer Venture Partners and Sorin Investments.
- AWS Preferred Partner building on Anthropic and Claude foundations.
Categories
About Aivar Innovations Private Limited
Aivar Innovations Private Limited is an AI-first technology partner that builds and operates data and ML platforms, generative-AI pipelines, and cloud-native infrastructure for enterprises. The company delivers consulting and engineering services—covering async Python platforms, multi-tenant SaaS architectures, compliance/PII safety layers, DevOps/SRE, and AWS modernization—often via reusable accelerators. Founded in 2024 and headquartered in Menlo Park, California, Aivar raised a $4.6M seed round in January 2026.