ArcaLabs
ArcaLabs

Forward Deployed Engineer

$180–280kPalo Alto, CAOn-siteMidFull-time

Posted Oct 20, 2026

The role in short

Deploys AI analyst agents inside investment banks and PE funds, building production workflows on FactSet data.

About the team

ArcaLabs builds AI agents that perform sell-side tasks (public comps, IC memos, pitchbook strip profiles, earnings comp analysis) end-to-end. We're partnered with some of the largest financial data providers, including FactSet and CapIQ. Our founding team previously built Arcafeed, a SaaS tool for buy-side analysts, which was acquired in late 2025. Our customers are the global heads of M&A, the MDs running pitches, and the analysts who used to spend their nights in Excel.

Forward Deployed Engineering is the team that turns research breakthroughs into production systems inside customer environments. We sit at the intersection of customer delivery, product, and core platform work, and we ship workflows that get used the same week we build them.

About the role

We're hiring a Forward Deployed Engineer (FDE) to lead end-to-end deployments of our agents inside investment banks, asset managers, and private capital firms. You'll fly to a customer's NYC, London, or Hong Kong desk on Monday, sit next to their analysts, and have a working agent running their Monday-morning comps refresh by Friday. Then you'll come back to Palo Alto and roll the patterns you discovered back into the core platform.

This is the highest-leverage engineering role at ArcaLabs. You'll build the workflows that define what a sell-side AI analyst can do, in the firm's house style, on FactSet's data, with full audit trails. You'll see the work used live within days.

Some of the work you'll do

  • Lead full-cycle deployments at top-tier banks and PE funds: discovery, scoping, building, customer training, and rollout.
  • Build agentic workflows on our own agent runtime, a forked coding-agent CLI that we run in E2B sandboxes and drive from a TypeScript control plane, wired to FactSet data, producing pitch-ready Excel, PowerPoint, and Word deliverables in each firm's house style.
  • Run agent-generated Python inside our E2B sandboxes, with persistent per-deal sessions and WebSocket bridges back to the control plane. Every credentialed tool call is brokered through Hardpan, our Rust secret vault and egress proxy, so the agent process never holds a customer API key.
  • Pair directly with MDs and analysts to model the workflows that matter: the comps templates, IC memo formats, and pitch strip profiles that are version-controlled in their muscle memory.
  • Write production Python and TypeScript against the controls a bank's security review actually asks about: a fail-closed audit trail, encrypted connector tokens, and per-sandbox secret brokering. You'll be the one answering those questions in the room.
  • Land the surfaces bankers already work in: per-environment manifests, Entra app registration, and tenant admin consent for the Excel, Word, and PowerPoint task pane and our Teams bot.
  • Identify reusable patterns across customers and champion them into the core platform with the rest of engineering.

You might be a fit if you

  • Have 3+ years of software engineering experience, including time on a customer-facing or platform-applied team (FDE, solutions engineering, or in-house engineering at a fund).
  • Are fluent in Python and TypeScript; have shipped agent-driven features in production: tool loops, sub-agents, streaming, and sandboxed code execution. We don't use an agent framework, so having built and debugged your own model loop counts for more here than any particular SDK.
  • Are comfortable in the Excel/PowerPoint/Word object models: openpyxl, python-pptx, and python-docx for writing, ExcelJS and JSZip for reading OOXML back. You've debugged a corrupt .xlsx, chased down an Excel repair-on-open prompt, or written formula provenance into cell comments.
  • Read a 10-K like a banker reads a pitch deck: fast, with an eye on the footnotes.
  • Are energized by sitting on a trading floor or in a deal room, not behind a Jira board.
  • Have opinions about agent design, evals, retrieval, and tool use, and want to put them into production this quarter.

Bonus points

  • Prior experience inside an investment bank, PE fund, asset manager, or financial data vendor (FactSet, Bloomberg, S&P, Refinitiv).
  • Have deployed an Office add-in or a Teams app into a locked-down enterprise tenant: manifest hosting, Entra registration, admin consent, and AppSource submission.
  • Terraform + AWS (RDS, S3, VPC). Terraform manages our data plane; services deploy from Dockerfiles and agent sandboxes are E2B templates defined in TypeScript.

Compensation

$180K – $280K + meaningful equity. Full benefits, relocation support, and daily breakfast, coffee, and snacks, plus complimentary fitness center access in Palo Alto. Travel covered to customer sites globally.

About ArcaLabs

ArcaLabs is the official FactSet AI Partner, building AI agents that run sell-side workflows for investment banks and PE funds. We're the team behind Arcafeed (acquired) and we're hiring the founding squad in Palo Alto. We believe AI will widen, not narrow, the set of people who can do high-end finance work. We want to ship that future.

Equal opportunity

ArcaLabs is an equal opportunity employer. We hire based on merit, and we accommodate applicants with disabilities. If you need accommodation during the hiring process, email careers@arcalabs.ai.

Work authorization

ArcaLabs does not sponsor employment visas or green cards, including new H-1B petitions and H-1B transfers. Applicants must be authorized to work in the United States without current or future employer sponsorship.

Prep for this interview

4 questions likely to come up at ArcaLabs, each matched to a line in this posting.

  1. 1Discovery and scoping
    They wrote: “Lead full-cycle deployments at top-tier banks and PE funds: discovery, scoping, building, customer training, and rollout.”

    A customer says “we want AI in our claims process.” What do you ask in the first meeting?

    How to answer

    Why they ask: Most FDE work starts this vague. They want to hear you find the real problem before you pick a tool.

    1. Map the current process: who touches a claim, in what order, with which systems
    2. Find the slowest or most error-prone step and ask what it costs them
    3. Agree on one measurable goal (for example time per claim, or error rate)
    4. Ask what data you can access, and who signs off on security

    Tip: Ask about the work, not the technology. Who does it today, how long it takes, what a mistake costs, and how they'd know it got better.

  2. 2Cloud, deployment and security
    They wrote: “Terraform + AWS (RDS, S3, VPC). Terraform manages our data plane; services deploy from Dockerfiles and agent sandboxes are E2B templates defined in TypeScript.”

    A bank wants our product inside their own cloud account, with no data leaving it. How do you approach it?

    How to answer

    Why they ask: Self-hosted and private-cloud deployments are common in regulated industries.

    1. Learn their rules early: network limits, approved services, review process
    2. Package the product so it installs the same way each time (containers, Terraform)
    3. Plan how you'll get logs and updates without touching their data
    4. Run a test install with them before the real one

    Tip: Talk about the conversation with their security team as much as the setup.

  3. 3LLMs, RAG and agents
    They wrote: “Have opinions about agent design, evals, retrieval, and tool use, and want to put them into production this quarter.”

    A customer wants a chatbot over their internal documents. How do you build it?

    How to answer

    Why they ask: RAG is the most common first project. They want the whole path, from messy files to answers people trust.

    1. Collect and clean the documents; handle permissions per user
    2. Split them into chunks, embed, and store with source links
    3. Retrieve, then answer with citations; say “I don't know” when retrieval is weak
    4. Build a test set of real questions with expected answers before launch

    Tip: Spend more time on the data and on testing than on the model.

  4. 4Coding
    They wrote: “Have 3+ years of software engineering experience, including time on a customer-facing or platform-applied team (FDE, solutions engineering, or in-house engineering at a fund).”

    Build a small in-memory cache. Then add expiry. Then a size limit.

    How to answer

    Why they ask: Staged exercises show how your code holds up when requirements change mid-interview, like at a customer.

    1. Start with get and set on a dictionary, plus tests
    2. Store an expiry time with each value and check it on read
    3. Track use order and evict the least recently used entry when full
    4. Say what you'd change for many threads or many machines

    Tip: Keep the first version simple and well tested so the next stage is a small change.

Not the one? Get the next one by email

New roles land here every day. Pick what you want to hear about.

FDE Info

Want the job, not just the posting? Learn how FDEs work.

FDE Info is our free weekly newsletter for forward deployed engineers: tools, field notes from real deployments, and how FDE teams hire and pay.

Free. One email a week. Unsubscribe in one click.

Similar roles at other companies

ArcaLabs

$180–280k · On-site

Apply