The signal: a sum of n terms where k appears only through k/n, and a factor 1/n, possibly hidden.
k=1∑nn+k1=n1k=1∑n1+k/n1→∫011+xdx=ln2.
Products: take logarithms. lnnnn!=n1∑k=1nlnnk→∫01lnxdx=−1, so nnn!→e1. Here the integral is improper. The step is justified because ln is increasing: the sum is squeezed between ∫1/n1lnxdx and ∫01lnxdx plus terms that tend to zero. In a written solution give this comparison, not just "Riemann".
The symmetry substitution
∫abf(x)dx=∫abf(a+b−x)dx.
If x↦a+b−x turns the integral into something that adds nicely to the original, I comes from an equation 2I=…:
∫0π/2sinnx+cosnxsinnxdx=4π: with x↦2π−x sine becomes cosine, and the sum of the two is ∫0π/21=2π.
∫0π1+cos2xxsinxdx=4π2: with x↦π−x the factor x becomes π−x, so 2I=π∫0π1+cos2xsinxdx=π⋅2π.
∫011+x2ln(1+x)dx=8πln2: first x=tant, then t↦4π−t, with 1+tan(4π−t)=1+tant2.
Cauchy–Schwarz for integrals
For f,g integrable on [a,b]:
(∫abfg)2≤∫abf2⋅∫abg2,
with equality (for continuous functions) only when one is a constant multiple of the other. The proof: the polynomial t↦∫(f+tg)2≥0 has discriminant ≤0.
The most common use is picking the "right" pair. For f>0 continuous on [0,1], with f and 1/f:
∫01f⋅∫01f1≥(∫011)2=1.
Pitfalls
A Riemann sum needs an integrable f. For improper integrals (like lnx at 0) you need a monotonicity or squeeze argument.
The symmetry substitution does not solve everything. If afterwards f(x)+f(a+b−x) does not simplify, look for another move.
For equality in Cauchy–Schwarz, state which functions are proportional; the grader wants to see why equality holds.
Further reading
Abbott, Understanding Analysis, 2nd ed., §7.2 (The Definition of the Riemann Integral)
Gelca & Andreescu, Putnam and Beyond, §3.2.9 (Riemann Sums)
Gelca & Andreescu, Putnam and Beyond, §3.2.8 (Definite Integrals)
Gelca & Andreescu, Putnam and Beyond, §2.1.3 (The Cauchy–Schwarz Inequality)
Gelca & Andreescu, Putnam and Beyond, §3.2.10 (Inequalities for Integrals)