New opportunity

Research Associate, BAC Agentic AI and Decentralized Finance

National University of Singapore · Singapore

About this role

The NUS Business Analytics Centre seeks a researcher for a project developing autonomous AI agents for decentralized finance (DeFi). Based at Kent Ridge Campus, the position works with faculty, industry collaborators and graduate researchers to design, implement and evaluate agents that reason, plan, interact with blockchain ecosystems and execute financial workflows securely. The posting is titled Research Associate, although its overview refers to the candidate as a Research Assistant.

Responsibilities

  • Design AI agent frameworks and autonomous or multi-agent systems for blockchain-based tasks. Integrate agents with Ethereum-compatible infrastructure and develop modules for smart contracts, wallets and DeFi protocols.
  • Implement Retrieval-Augmented Generation (RAG), memory systems, tool calling and agent orchestration. Review research on AI agents, large language models, blockchain and DeFi.
  • Benchmark and experimentally evaluate agent performance, including security testing. Translate industry partners’ requirements into research prototypes; prepare technical reports, publications and documentation; and support demonstrations and technology transfer.

Required qualifications

A master’s degree in a relevant field; experience with modern AI development frameworks; good knowledge of blockchain, smart contracts and DeFi; proven experience delivering industry-facing or commercial AI solutions; and strong communication skills, including with non-technical stakeholders. No specific number of years of experience is stated.

Beneficial experience includes agent orchestration, function calling, workflow automation, planning and reasoning algorithms, RAG and Model Context Protocol; Ethereum, Solidity, Hardhat or Foundry, Web3.py, Ethers.js and wallet integration; and DeFi applications such as automated market makers, decentralized exchanges, lending, yield farming, staking, cross-chain protocols, liquidity pools and tokenomics.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

AI agentsMulti-agent systemsAgent orchestrationTool callingFunction callingWorkflow automationPlanning and reasoning algorithmsLarge Language ModelsRetrieval-Augmented GenerationModel Context ProtocolBlockchainWeb3EthereumSolidityHardhatFoundryWeb3.pyEthers.jsWallet integrationSmart contractsDecentralized FinanceAutomated Market MakersDecentralized ExchangesLending protocolsYield farmingStakingCross-chain protocolsLiquidity poolsTokenomicsAI security

Your skill match

Checking your profile…

YOUR NEXT STEP

Get interview-ready for this role

A focused preparation guide, built around this job’s responsibilities and requirements.

✦ AI-generated guide
Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.
01

Understand the role

What matters most

  • Build agents that execute blockchain tasks

    The role calls for autonomous and multi-agent systems that reason, plan and execute blockchain-based tasks, with integrations to Ethereum-compatible infrastructure, smart contracts, wallets and DeFi protocols.

  • Implement reliable agent workflows

    The posting specifically calls for RAG, memory systems, tool calling and agent orchestration. Prepare to explain how these components support an end-to-end financial workflow, rather than discussing them as isolated techniques.

  • Evaluate performance and security

    Benchmarking, experimentation and security testing are explicit duties. Prepare to describe how an agent would be tested for task completion and unsafe behavior when interacting with blockchain systems.

  • Turn partner needs into research outputs

    The researcher works with industry collaborators to translate business requirements into prototypes, while also preparing reports, publications, documentation and demonstrations. Clear communication with non-technical stakeholders is a stated requirement.

Likely challenges to think through

  • Balancing autonomy with transaction safety

    Preparation hypothesis: An agent that can execute financial workflows may need enough autonomy to complete tasks without taking unintended actions. Prepare an example design that limits tool or wallet actions and tests behavior before execution; this is a reasoned trade-off, not a stated project problem.

  • Producing meaningful evaluation results

    Preparation hypothesis: Comparing agent approaches may be difficult if successful task completion, reliability and security are measured differently. Prepare a proposed benchmark with representative blockchain tasks, explicit success criteria and security tests. The posting requires evaluation but does not prescribe metrics.

  • Reconciling research and partner deliverables

    Preparation hypothesis: A publishable investigation and an industry-facing prototype may favor different levels of novelty, robustness and speed. Prepare to discuss how you would define a narrow research question while documenting a demonstrable prototype and communicating its limits.

03

Practice your answers

Technical practice

How would you design an agent that plans and executes a blockchain-based financial workflow?

What to coverThe posting calls for agents that reason, plan and execute blockchain tasks securely. Suggested preparation: explain how you would represent goals and state, check a plan before execution, and decide when the agent should stop or request review.

When would you use multiple agents rather than one agent for a Web3 task?

What to coverThe role includes autonomous and multi-agent systems, orchestration and tool calling. Suggested preparation: compare a single-agent design with specialised agents, including coordination costs, tool permissions and ways to detect failed handoffs.

How would you use RAG and memory in a DeFi agent?

What to coverThe posting explicitly names RAG and memory systems. Suggested preparation: distinguish retrieved reference material from persistent task state, and explain how you would check source relevance and prevent stale information from driving an action.

How would an agent interact with an Ethereum-compatible network and a wallet?

What to coverThe duties include blockchain integration and modules for wallets and DeFi protocols. Suggested preparation: walk through reading network state, preparing a transaction, obtaining authorisation and checking its outcome; distinguish what the agent may propose from what it may execute.

How would you build and test a module that interacts with a smart contract?

What to coverSmart-contract interaction is a stated duty; Solidity, Hardhat, Foundry, Web3.py and Ethers.js are listed as beneficial experience, not mandatory tools. Suggested preparation: discuss contract interfaces, input validation, transaction failures and a testing approach using tools you actually know.

What should an agent check before carrying out an AMM swap or lending action?

What to coverAMMs, decentralised exchanges and lending protocols appear in the posting's beneficial-experience list. Suggested preparation: use one of these as an example to discuss protocol state, expected outcome, transaction costs and conditions under which the agent should decline to act.

How would you security-test an agent that can initiate DeFi actions?

What to coverThe project aims for secure execution, and security testing is an explicit responsibility. Suggested preparation: describe tests for unsafe tool calls, misleading inputs, excessive permissions and unexpected transaction outcomes, including how failures would be recorded.

How would you benchmark whether a DeFi agent performs well?

What to coverThe role includes benchmarking and experimental evaluation of agent performance. Suggested preparation: propose representative tasks, baselines and measures for task completion, correctness, efficiency and safety; explain how you would make comparisons reproducible.

How would you turn a research paper on agents or DeFi into a testable prototype idea?

What to coverThe duties include literature reviews across AI agents, LLMs, blockchain and DeFi, alongside research and system development. Suggested preparation: explain how you would identify a paper's assumptions, select a claim worth testing and design an experiment relevant to the project's agent workflows.

Behavioural practice

Tell me about a time you translated an external partner's needs into a technical prototype.

What to coverThe posting calls for translating industry partners' requirements into research prototypes. For this suggested practice question, use a real situation, the actions you took to clarify priorities and scope, and the resulting prototype or decision; do not assume the partner's first request was the final specification.

Describe a time you explained a complex AI or technical risk to a non-technical stakeholder.

What to coverStrong communication with non-technical stakeholders is a stated qualification. For this suggested practice question, give the situation, how you adapted the explanation or checked understanding, and what decision or outcome followed.

Tell me about a research or development task you coordinated with people from different disciplines.

What to coverThe role works with faculty, industry collaborators and graduate researchers. For this suggested practice question, describe a real situation, your contribution to aligning responsibilities or resolving a disagreement, and the result.

Describe a time testing changed your approach to an AI system.

What to coverThe posting pairs system development with experimentation and security testing. For this suggested practice question, explain the situation, what the evidence revealed, the action you took and the measured or observed result.

Tell me about a time you documented or demonstrated a technical result for others to use.

What to coverTechnical reports, publications, documentation, demonstrations and technology transfer are listed duties. For this suggested practice question, give the situation, how you chose what to communicate, and the result for the intended audience.

02

Sharpen your knowledge

Topics to prepare

  • Agent planning and multi-agent coordination

    Review how an agent selects actions, delegates tasks and handles failed steps; sketch an agent workflow for a blockchain-based task. The posting explicitly calls for autonomous and multi-agent systems that plan and reason.

  • Blockchain and smart-contract fundamentals

    Trace how a wallet signs a transaction and how a contract call changes state; explain where an Ethereum-compatible integration can fail. Knowledge of blockchain and smart contracts is a stated qualification, and these integrations are part of the duties.

  • DeFi protocol mechanics

    Compare an automated market maker trade with a lending transaction, identifying inputs, state changes and financial risks. DeFi knowledge is required; AMMs and lending are listed as beneficial areas of experience.

  • Agent reliability and security

    Outline failure modes for an agent executing a financial workflow, including incorrect tool calls and unsafe contract interactions, then propose checks. Secure execution, benchmarking and security testing are explicit project duties.

  • RAG, memory and tool-use design

    Diagram when an agent should retrieve information, retain state or call a tool, and identify one failure mode for each. Implementing RAG, memory systems, tool calling and orchestration is explicitly listed.

Skills to revise

  • Prototype an agent workflow

    As a preparation exercise, implement a small agent that plans steps, calls a tool and records outcomes; explain how you would test failed or repeated calls. The role involves building agent frameworks and workflows; no particular AI framework is specified.

  • Build a contract-interaction module

    As a preparation exercise, use a posting-listed option such as Web3.py or Ethers.js to prepare a wallet-connected contract call in a safe test setting, and document transaction checks. Smart-contract and wallet modules are explicit duties; those libraries are listed as beneficial, not required.

  • Implement a bounded RAG workflow

    As a preparation exercise, build retrieval over a small set of DeFi documentation, add source-aware responses and test an irrelevant retrieval result. RAG implementation is an explicit duty; the particular data set is only a practice suggestion.

  • Design a reproducible evaluation

    As a preparation exercise, create test cases for task completion, erroneous actions and security failures; record metrics and compare agent runs. The posting calls for benchmarking, experimentation and security testing.

  • Translate requirements into a research prototype

    As a preparation exercise, turn a hypothetical non-technical request for a DeFi workflow into a brief specification, prototype scope and demonstration plan. The role involves working with industry partners, communicating with non-technical stakeholders, documenting results and supporting demonstrations.

04

Your preparation checklist

  • Map the stated qualifications to evidence

    Prepare a one-page list of evidence you can discuss for a relevant master’s degree, modern AI development frameworks, blockchain, smart contracts, DeFi, and communication with non-technical stakeholders. These are stated qualifications; do not claim experience you cannot substantiate.

  • Prepare an industry-facing AI project walkthrough

    Select a delivered commercial or industry-facing AI solution, if you have one. Be ready to explain the stakeholder need, your contribution, technical decisions, outcome, and how you communicated trade-offs; delivery experience is a stated qualification.

  • Sketch an agent architecture

    As a preparation exercise, draw an autonomous or multi-agent workflow that plans a blockchain task. Show where RAG, memory, tool calling, and orchestration would fit, and identify which actions require safeguards. These components appear in the duties.

  • Work through a DeFi integration example

    Prepare a small design or existing-code walkthrough for an agent interacting with a wallet, smart contract, and one DeFi protocol on Ethereum-compatible infrastructure. Explain transaction approval and failure handling as proposed design choices, not as a specified NUS architecture.

  • Draft an evaluation and security plan

    Create a short table of proposed agent benchmarks, experimental comparisons, and security tests, including how you would assess task success and unsafe actions. The role explicitly includes benchmarking, experimentation, and security testing.

  • Summarize relevant research and documentation

    Bring a concise literature summary spanning AI agents or LLMs and blockchain or DeFi, with one open question worth testing. Prepare a sample technical report outline to discuss how results could become documentation or a publication.

  • Prepare collaboration and scope questions

    Write questions about industry partners’ prototype needs, expected demonstrations and research outputs, and how faculty and graduate researchers collaborate. Ask the employer to clarify the Research Associate title versus the overview’s Research Assistant wording.

AI-generated from the job information · Prepared Sep 30, 2026 (UTC)Useful preparation, not a guarantee of eligibility or an offer.

Availability last checked: 4 Nov 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in Singapore. Matched to this role, not your profile.

Explore more jobs
Finding similar opportunities…