Example · Credit Analyst
65% of your week as a credit analyst is still yours.
Software now spreads financials and runs the ratios. Setting the grade and presenting it stays with you.
Make my planA typical week at the middle pace. Your plan uses your own tasks.
Is credit analyst a good career with AI?
In a typical week at the middle pace, 38% of this work needs your judgment and 27% is work you do using AI. AI could do about 35%. Spreading, ratios, comps and document chasing are 34% of your week, and a model plus your bank's own vendors can already do all of it. That is the part most analysts dread.
Read more
It is also what justified hiring a new graduate, and what a spreading module is sold to remove. Nobody is coming for the catch, the risk rating, or the defense at committee.
Most exposure analyses miss one point. Lending is regulated, but the rules name the bank and the approver. They do not name the analyst. Your protection comes from whoever signs, and that protection can be handed to a vendor. This year the federal model-risk guidance was rewritten to apply mainly to banks over $30 billion (Sullivan & Cromwell, April 2026), and in May 2025 the CFPB, the federal consumer finance regulator, withdrew its guidance on explaining algorithmic declines. Your protection exists, but it is thinner than "regulated" suggests, and it is tied to a seat you do not yet hold. This plan aims you at that seat.
The week, task by task
Who does each hour of a typical 40-hour week.
See every task
Needs your judgment.
- Risks, mitigants and structuring recommendation8%
- Presenting and defending the credit at committee8%
- Borrower and RM interaction: site visits, calls, negotiating what credit will accept7%
- Risk rating assignment and grade justification5%
- + 2 smaller tasks10%
AI does the slow parts. You still own it.
- Credit memo drafting: borrower background, financial analysis narrative, collateral summary11%
- Covenant monitoring and annual reviews: compliance testing, tickler, annual write-ups10%
- Collateral analysis: appraisal review, rent rolls, lease abstracts, stressed LTV and cap rates7%
AI could do most of it.
- Spreading financial statements and tax returns into the spreading tool16%
- Ratio generation, DSCR, global cash flow, sensitivity tables7%
- Industry and market research, peer comparisons, CRE market write-ups6%
- Document collection: chasing statements, checking completeness, file assembly5%
Why some tasks still need a person
Finding what the relationship manager didn't mention
This takes context. The related-party lease isn't labeled "related party." Knowing the guarantor's brother owns the landlord entity comes from knowing the bank's files and one conversation. A model reads what is in the package, and the catch is about what is missing.
Writing the risks, mitigants and structure recommendation
Your accountability here is borrowed, but it sits close to the signature. This is the part of the memo the approver relies on and examiners read when a loan goes bad. Anyone can get a draft. A recommendation someone will sign is worth much more.
Assigning and defending the risk rating
This needs accountability and checking. A wrong grade surfaces at the next exam or the next downgrade, and the bank's allowance depends on it. Someone with a name must own the grade.
Presenting at committee
This needs you in the room and answering for the credit. The chief credit officer's question tests whether a human understood the credit. It is the task many analysts dread, and the single most protected one. The plan changes who owns the questions.
Watch-list and problem-loan work
This takes context and accountability. You make early-warning calls on a credit the bank already holds, working with a workout attorney and an RM (relationship manager) who wants to keep the relationship. None of it fits a template.
Good to know
What is hard about this field, and what you can do about it.
The hard truth
More business lending now happens outside banks.
By mid-2024, US private credit had grown to about $1.3 trillion, roughly five times its 2009 size. BLS expects credit analyst jobs to shrink.
Sources (2)
- US private credit totaled $1.34 trillion by mid-2024, roughly five times its 2009 size. Federal Reserve Board, FEDS Notes, 2025-05
- Credit analyst jobs are projected to decline 4.3% from 2025 to 2035. BLS Employment Projections, 2026
What's next for a credit analyst?
Plan A is the best move inside the work you do now.
Grow in the job you have
The main focus. Most weekly sessions build toward it.
- Each week
- 4 sessions
- Ready in
- About 9 to 15 months
Why it fits
Why you. Your memo history is a track record the bank can audit, and the next step is signing something. Loan policy already defines the smallest authority tier. The open question is whose name goes in it.
The gap. You have no proof of your judgment separate from your manager's, no record of carrying a credit alone at committee and no credential. The 13-week plan covers all three.
The trade-off. Your name goes on losses. The number of people with authority is small and set by policy, so you may hear "when a seat opens." It is also tied to your bank. If the bank is bought, it resets, which is why week 10 looks at other banks.
2 other moves that fit
- Portfolio management and problem loansReady in about 6 to 9 months
- Same seat, larger bankReady in a few weeks of applying once you hold the crc
What's a plan B for credit analysts?
Plan B is meant to be different from this job. It was picked because skills from this role carry over.
Get ready for a new kind of job
A second option. A newer job AI is creating that these skills already fit.
Finding what the RM did not mention is the catch banks need when AI starts spreading financials and drafting memos.
- Each week
- 1 session
- Outlook
- Hiring now
- Pay
- $230K to $270K a year (one posting)One Anthropic safety evaluations posting, 2026; AI lab pay runs well above most employers.
2 other options
- AI governance analystDefending a risk rating and documenting exceptions are close cousins of model risk review, which some banks now extend to AI tools.
- Expert AI trainer (contract)Writing risks, mitigants and structure recommendations is expert work that AI training platforms screen for and pay well on contract.
In your own plan you can switch to any of the 47 roles.
The 13 weeks
Small steps that add up. One plan for both: most sessions build Plan A, one a week builds Plan B.
By week 13 you will have
AFor Plan A
- A ledger of your recommendations against outcomes
- A plain-language decline reasons template
- A submitted CRC application
BAnd for Plan B
- A draft write-up of your AI test
- A published write-up of your AI test for your portfolio
- 1Measure your week
- 2Test AI on spreading
- 3Track committee questions
- 4Own the covenant rules
You finish withYour manager's answer on owning the exception rules
- 5Map the rules that apply
- 6Improve decline reasons
- 7Learn watch-list work
- 8Check your CRC readiness
You finish withA practice score on your weakest topic
- 9Present a loan yourself
- 10Check other banks
- 11Gather your results
- 12Make your ask
- 13Decide your next step
You finish withA one-paragraph decision with a date
Three courses from the plan
- Credit Risk Certification (CRC)ProSight Financial Association (RMA merged with BAI to form ProSight, November 2024; the old rmahq.org/credentialing page now 404s) · $550 initial; $250 retake; $150 recertification every 3 years (45 CE credits)
Yes. This is the credential in the plan.
- Commercial Lending School – FoundationalAmerican Bankers Association · ABA members $2,995; nonmembers $3,995
Not for you. It is pitched at "junior-level" lenders. Ask about Advanced only on bank money.
- Commercial Banking & Credit Analyst (CBCA)Corporate Finance Institute · $347.90/yr self-study (30% off $497); $592.90/yr full-immersion
No, per the advice we decline. Listed so you can check us.
Checked 2026-09-22.
The fuller picture
Risks beyond AI: highMerger risk
Banks keep consolidating, and an acquisition collapses two credit departments into one and resets every lending authority in the acquired one, with no tool involved. If your bank is a plausible target, that matters as much as anything in the AI analysis.
Where this could be wrong
- Bank size drives the regulatory read. We assumed a bank under $30 billion. Above $30B, SR 26-2 applies in full, a model-risk function exists, and the "learn Python" advice we declined becomes a real internal path.
- Acquisition may dominate everything. If your bank is bought, the authority resets and "same seat, larger bank" becomes the move to lead with. Week 10 is the hedge, but it may not be enough.
- The tool may be better at add-backs than we assumed. Then the Week 2 list is short and your claim on the exception rules is weaker.
- Regulation could tighten. If states follow Colorado's January 2027 human-review rule, this report understates your protection.
Also matched: credit analyst, commercial credit analyst, credit underwriter. Research checked 2026-09-22. Page updated .