Fiska is transforming how finance teams manage event spend. We're building the first AI-powered platform that automates expense tracking, budget management, and real-time reconciliation for social media and marketing professionals (SMMPs).

Our customers save an average of $250K+ annually and reduce reconciliation time by 78%. We're backed by leading investors and working with global brands to solve billion-dollar problems in event finance.

As a Founding AI Engineer at Fiska, you'll be at the cutting edge of building production-grade AI systems that power our intelligent expense management platform. You'll own the entire AI stack—from research and experimentation to production deployment—and directly impact how thousands of finance teams interact with financial data.

This is a rare opportunity to shape the AI/ML direction of an early-stage startup, working directly with cofounders to build agentic systems, implement RAG pipelines, and create AI experiences that feel magical to users while maintaining enterprise-grade reliability and security.

Take agentic AI SMMP workflows from concept to production, owning the complete pipeline including retrieval, reasoning, tool integration, and safety measures while continuously improving accuracy through data-driven iteration

Evaluation Infrastructure: Architect robust evaluation frameworks for prompts and models with both offline and online testing capabilities, enabling rapid experimentation through A/B testing and systematic performance comparison

Build sophisticated context management systems incorporating RAG, long-term memory, intelligent summarization, and data compression to enhance both conversations and automated expense categorization

Internal Tools: Develop reusable tools and SDKs that empower engineering teams and AI systems to build faster, creating a multiplier effect across the organization

Manage integrations across multiple LLM providers with intelligent fallback mechanisms, balancing cost efficiency, response speed, and output quality for optimal performance

Safety & Compliance: Design and deploy robust guardrails,



and implement safety protocols ensuring AI-generated outputs meet stringent financial industry compliance and regulatory requirements

Establish clear success metrics covering accuracy, latency, cost, and safety, design and execute A/B tests, and integrate user feedback loops to drive measurable quality improvements

Proactive and self-motivated engineer with strong ownership mentality and ability to drive projects independently from inception to completion

Builder's mindset with comfort navigating ambiguity, rapid prototyping skills, systematic debugging approach, and relentless focus on user experience

Demonstrated expertise in large language models through professional work or exceptional personal projects showcased on your resume

Proven experience integrating multiple LLM providers with tool-calling and function-calling capabilities, comfortable switching between models and implementing robust fallback strategies

Deep hands-on experience with RAG systems including indexing strategies, chunking techniques, embeddings, and reranking, with understanding of reliability and cost-latency optimization

Solid computer science fundamentals including data structures, algorithms, comprehensive testing practices, and performance profiling

Advanced prompt engineering skills with experience building prompt libraries, analyzing failure modes, and systematically refining prompts and policies

Strong analytical ability to establish meaningful KPIs covering accuracy, latency, cost, and safety, with experience designing A/B tests and implementing human feedback systems

Production experience with MongoDB including schema design, optimization, and scaling





Proficiency in Python with hands-on experience in modern ML/AI frameworks and libraries

Working knowledge of cloud-native vector databases such as AWS OpenSearch with k-NN, Amazon MemoryDB for Redis, GCP Vertex AI Vector Search, pgvector on Aurora/AlloyDB, or managed services like Pinecone, Qdrant, or Weaviate

Experience with AWS Bedrock Agents, GCP Vertex AI Agent Builder, or modern agent frameworks like LangGraph, CrewAI, and AutoGen for building production-grade agentic systems

Hands-on experience with cloud AI platforms including AWS Bedrock, GCP Vertex AI, or Azure OpenAI for model deployment and orchestration

Proficiency with LLM observability and evaluation platforms such as AWS CloudWatch with Bedrock integration, GCP Cloud Trace, Vertex AI Model Monitoring, Langfuse, Weights & Biases, or MLflow

Experience with cloud-based fine-tuning and model optimization using AWS SageMaker, Amazon Bedrock Custom Models, GCP Vertex AI Training, or Model Garden

Advanced understanding of retrieval systems, embeddings architecture, safety mechanisms, guardrail implementation, and red-team testing methodologies

Academic or industry background in natural language processing or machine learning research with publications or patents

Competitive salary aligned with top Indian startups plus significant equity stake to share in our growth

Health Insurance

Top-tier health insurance covering you and your family with major hospitals across India

Remote First

Work from anywhere in India with flexible hours and home office setup allowance

Adaptable Time Off

Generous leave policy with paid time off, sick leave, and all Indian national holidays

Annual learning allowance for AI/ML courses, certifications, research papers, and tech conferences

Career Growth

Early-stage startup with rapid growth opportunities and direct impact on company direction

Fast & Agile

Ship quickly, iterate constantly, and adapt to change.

We build for the long term—sustainable growth, sustainable practices, sustainable impact.

We invest in our communities and create opportunities for all.

📌 AI Programmer - Remote Working (Itanagar)
🏢 fiska.AI
📍 Itanagar